From: IN%"Per.Jensen@hhyg.slu.se" 18-APR-1995 01:28:57.44 To: IN%"applied-ethology@sask.usask.ca" CC: Subj: Weaving in horses Dear all, I am slightly involved in a court case: A riding school bought a seven-year old horse. After some time they contacted the seller and wanted the seller to take the horse back, because it was weaving. The buyers claims that this is a deficiency of the horse. The seller claims that the horse had never weaved before. It was kept as a family horse, used to pasture. They refuse to take the horse back. I have little experience and knowledge of weaving, but based on the assumption that this is a typical stereotypy, my opinion is that this can never be a deficiency in the individual horse in the sense that there are some horses that are weavers and others that are not. A stereotypy develops as a result of an interaction between an animal and its environment. An individual horse may be more or less inclined to develop the behaviour. However, given that it can be proven that this horse had not shown weaving before arrival to the riding school, I think it must have developed there as a result of some environmental deficiency relevant for this particular individual. I would welcome any additional opinions on this subject in general and specifically on the case. It is also claimed by the buyers that this behaviour is 'contagious' and will spread to the other horses. In my experience from numerous public lectures, nowhere is the idea of the 'contagiousness' of stereotypies more wide-spread than among horse-people. The only study I know of that has actually looked into this is the one on sows by Alistair Lawrence and colleagues and it did not support the idea. Is there anyone who has any hard data on the phenomenon regarding horses? Looking forward to your reactions! Per ******************************************************************* Per Jensen Professor of Ethology __/\______________9 Swedish University of Agricultural Sciences,\ o I Department of Animal Hygiene, \- I Section of Ethology \_______________I SKARA, SWEDEN /\ /\ E-mail: Per.Jensen@hhyg.slu.se / \ / \ ******************************************************************* From: IN%"Per.Jensen@hhyg.slu.se" 18-APR-1995 03:04:34.97 To: IN%"applied-ethology@sask.usask.ca" CC: Subj: n Jeff Rushen wrote: > >Mike is correct. Regrettably the experiment with the lions would not be >possible: Per Jensen cannot be replicated. Sorry to dissapoint you, but of course I can replicate myself. Ask my four kids! However, they were reluctant to take part in the experiment. They could consider taming one of our bantam cockerels, though, if you like. Three groups of 10 cows in treatment A and three groups of 10 >cows in treatment B gives n=3 (the number of groups) not n=30 (the number >of cows). Where 36 piglets from three sows are used in treatment A, while >36 piglets from another three sows are used in treatment B, (even if the >piglets are housed individually) the "n" is not 36 (the total number of >piglets) but 3 (the number of litters). This is because the genetic >relatedness of litter mates means that litter mates cannot be considered as >independent replicates. > > I disagree! Clearly this must be a question of what you are interested in? I have just finished a study involving 250 piglets. I wanted to study aggressive behaviour when the piglets were mixed in groups under different conditions. I mixed them in groups of about 50 piglets. For each piglet, I have recorded a number of variables, including litter and treatment. The treatment of each piglets involves being mixed in a group with a particular characteristic. Clearly, I can use the pig as observation unit and in my analysis check for effects of genetics, etc. Each series of different groups would constitute a replicate. Per PS. Mike, don't send my manuscript to Jeff Rushen for reviewing, please. ******************************************************************* Per Jensen Professor of Ethology __/\______________9 Swedish University of Agricultural Sciences,\ o I Department of Animal Hygiene, \- I Section of Ethology \_______________I SKARA, SWEDEN /\ /\ E-mail: Per.Jensen@hhyg.slu.se / \ / \ ******************************************************************* From: IN%"Per.Jensen@hhyg.slu.se" 18-APR-1995 03:05:40.08 To: IN%"applied-ethology@sask.usask.ca" CC: Subj: Causes and functions Howdee, Thanks for contributions to this issue. Some of you have done good jobs to present the base of the discussion. However, my main concern was not that I (or others) did not know the differences between causal and functional explanations. My problem is that we all know it, but it is so easy to forget about it once we have to formulate our specific hypotheses and questions and design our experiments. Why do animals fight with strangers? Anyone skimming the literature will find that it is often claimed that they do so to establish dominance. This is a classical example of mixing explanatory levels. Establishing dominance may be the consequence of the behaviour and even the function of it but it can never be its cause. The cause has to come before the effect. The cause must be explained in terms of what stimuli that elicit the behaivour. In pet therapy, you frequently see aggressive behaviour classified in different groups, requiring different treatments. Often, one group may be for example fear-induced aggression, another dominance aggression. But these are different levels, any aggressive behaviour may belong to both categories. How can you treat animals if the diagnosis mixes levels of explanations? Another area where levels are frequently mixed is that of parent-offspring conflict. Trivers showed good evidence that we should expect a conflict between parents and young over the amount of investment. People have searched for such conflicts in the overt behaviour of the animals. However, as most people know, the theory speaks about an ultimate conflict. The proximate resolution may not look like a 'conflict' at all. Again, grab the journals and skim them: When we don't find any obvious signs of conflict, we blaim the fact that the theory is an ultimate one after all, so the data do not contradict them. When we do find signs of conflict, the data suddenly DO support the theory. Again, mixing levels causes only confusion. However, a case which does tend to complicate the issue is that of goal orientation, and here I don't differentiate between the categories nicely outlined by Jonathan Cooper (although Jon is of course quite right). A goal which sets the endpoint or the sollwert of a behaviour always relates to the future and in that sense will always be closer to the function than the causes eliciting the behaviour. I will argue that the Tinbergen four questions maybe are not complete, because a behaviour is not only turned on (causation) but is also turned off. This is where goal orientation comes in. Tinbergen obviously did not think about this, but I think we may need a fifth level, the switch-off mechanism of a behaviour. Mixing that level with the turn-on level may be equally fallacious. For example, Mike Baxter, more than ten years ago suggested that sows may not have to engage in nest building behaviour at all if you only provide her with the goal for the behaviour, i e a nice, comfortable nest. Experiments have failed to support this. Why? Because what starts the behaviour is different from whatever stops it. Nesting is started by prolactin increase but what makes it stop is an open question. Similarly, fighting between unacquainted pigs is started by something like stimuli asociated with strangeness. However, it stops because of different things. What about that? By the way Jeff - when you work hard to find an angle on how to disagree completely with this message, remember that I may know one or two things about your private life that you probably would hate to see in public, even if only in cyberspace............. Cheers all ******************************************************************* Per Jensen Professor of Ethology __/\______________9 Swedish University of Agricultural Sciences,\ o I Department of Animal Hygiene, \- I Section of Ethology \_______________I SKARA, SWEDEN /\ /\ E-mail: Per.Jensen@hhyg.slu.se / \ / \ ******************************************************************* From: IN%"ALund@ZI.KU.DK" "Lund, Anders {ZI}" 18-APR-1995 05:53:56.83 To: IN%"applied-ethology@sask.usask.ca" "ISAE-net" CC: Subj: RE: Coprofagous dogs The question was raised if for instance the protein level in cat faeces could motivate dogs to eat it. By putting the question to a nutrionist I got the following answer: It seems that domestication has reduces cats' ability to digest proteins, therefore cat food is much higher in protein content than dog food. (The difference in protein content in canned cat and dog food is so great, that a cat fed with only dog food would get metabolic problems (and possibly psychic ones?) whereas a dog fed on exclusively cat food would get liver problems). Furthermore, pet cats usually get more food than they need. These factors make the cat faeces high in protein and could therefore motivate the dog to eat it. However, there are as already mentioned numerous reports on dogs eating faeces from pups and humans (particularly children), so it coul also be a behaviour based on keeping the home range clean from olfactory giveaways. In the concrete situation between a domestic dog and cat there could also be a dominance situation: removal of teritorial markers. Finally, dominant cats would put their faeces as high as possible, as markers, and the continuous removal of these markers may (actually) cause psysic disturbances in the cat and cause it to become more aggressive. Yours Anders Lund From: IN%"peter.penning@bbsrc.ac.uk" "PENNING" 18-APR-1995 06:45:12.09 To: IN%"applied-ethology@sask.usask.ca" "applied ethology" CC: Subj: Dogs eating faeces I have noticed on many occasions that dogs will also eat dung from ruminants and horses. It was suggested that they may eat cat faeces because of the protein content, what then is the basis for eating faeces from herbivores? Any ideas. Peter Penning From: IN%"RUSHENJ@NCCCOT.AGR.CA" "JEFF RUSHEN" 18-APR-1995 07:14:48.31 To: IN%"APPLIED-ETHOLOGY@sask.usask.ca" CC: Subj: Causes, functions Per writes: >However, my main concern was not that I (or others) did not know the >differences between causal and functional explanations. My problem is that we >all know it, but then he discusses a causal explanation as an example of a functional explanation. >Why do animals fight with strangers? Anyone skimming the literature will >find that it is often claimed that they do so to establish dominance. This >is a classical example of mixing explanatory levels. Establishing dominance >may be the consequence of the behaviour and even the function of it but it >can never be its cause. The cause has to come before the effect. The cause >must be explained in terms of what stimuli that elicit the behaivour. Is this really a functional explanation i.e. one that explains the behaviour by its consequences? It is more likely a causal one that explains the fight by postulating a desire of the pigs to become dominant, and an expectation that they will be dominant if they fight. These hypothesized "desires" and "expectations" exist before the fight and so this is a causal explanation. Note that this explanation may be difficult to test, or it may be wrong. But it is still a causal explanation. Such purposive explanations of animal behaviour have been replaced by functional ones in many cases i.e. natural selection has resulted in pigs that behave AS IF they were fighting with the purpose to become dominant. This is the "goal-seeking" or "goal-attaining" behaviour discussed by McFarland (1989) (Problems of Animal Behaviour), who distinguishes it from goal-directed behaviour (where the animals really do have a cognitive representation of their goals). >What about that? By the way Jeff - when you work hard to find an angle on >how to disagree completely with this message, remember that I may know one >or two things about your private life that you probably would hate to see >in public, even if only in cyberspace............. So the gloves are off. No more Mr. Nice Guy! I have some inside information on Per's recent attempts to sell a horse if any one is interested! Jeff Rushen From: IN%"Chris.Sherwin@bristol.ac.uk" "CM. Sherwin" 18-APR-1995 08:06:50.47 To: IN%"applied-ethology@sask.usask.ca" CC: Subj: weaving in horses Dear Per You might want to contact Jonathon Cooper or Christine Nicol regarding learning ("contagiousness") of stereotypies. Their work was using voles, but they are possibly a suitable model. Chris Sherwin From: IN%"mjk1000@cus.cam.ac.uk" 18-APR-1995 08:14:10.19 To: IN%"applied-ethology@sask.usask.ca" CC: Subj: RE: Weaving in horses Dear Per (and others), I have some experience with the study of stereotypies in horses and with horses and horse people (as a riding instructor before returning to academia), so here's my two pence worth! I agree with Per's interpretation of the development of the stereotypy in this case- the weaving stereotypy has been suggested to be related to motivation to perform locomotory behaviour, which may be thwarted by placing a horse in a confined environment. It would be very interesting to know more about the housing at the riding school- whether the horse was housed in a loosebox or in a stall, etc. If this particular horse was housed at pasture prior to moving to a more confined environment at the riding school (for instance, housed in a loosebox at the riding school and rarely turned out to pasture), and did not perform the stereotypy before moving, then it would indeed seem to be the case that the stereotypy developed in response to the confined environment at the riding school. As Per states in his original posting, the stereotypy is not a deficiency in the horse, but due to an interaction between the animal and its environment. As for the weaving stereotypy being contagious, and the common belief held by horse people that horses adjacent to a weaver will begin weaving by following its example, I am not aware of anything other that anecdotal evidence to support this belief. It should be noted that the development of stereotypies by horses housed adjacent to each other may not be due to simple mimicry of the original stereotyper, but due to the fact that all the animals are experiencing the same adverse environment and develop stereotypies in an attempt to cope with it. It is possible that the coping METHOD may be copied from the neighbour- i.e. a horse experiencing adverse conditions mimics the neighbour's behaviour initially, finds this provides some degree of relief from these conditions, and thus the stereotypy develops. I look forward to hearing the thoughts of others on this very interesting and important topic. Best Wishes, Mark J. Kennedy, University of Cambridge, Dept. of Clinical Veterinary Medicine, Madingley Road, Cambridge CB3 OES, U.K. (mjk1000@cus.cam.ac.uk) From: IN%"RUSHENJ@NCCCOT.AGR.CA" "JEFF RUSHEN" 18-APR-1995 08:25:56.00 To: IN%"APPLIED-ETHOLOGY@sask.usask.ca" CC: Subj: "N" From Chris Sherwin: > Firstly, isn't the principle behind statistical analysis that the >tests are devised to cope with variable numbers, and that the probability >of a significant finding is dependent on N. Therefore, if the P value is >less than 0.05, it is significant whether 2, 20 or 200 animals were used. >I accept that significant findings based on N=600 are likely to be more >generalisable than those based on N=6, but if this is the concern of >referees they should state so. They should not imply that the statistics >are "wrong". Strictly speaking, yes. However, most critical values in statistical tests are based on the assumption that the animals in the sample used were drawn at random from the population that you wish to generalize the results to. This is never done, but it is more obviously never done with an n=2 than an n=200. A group of 200 is less likely to contain animals which are all brothers, all crazy, or all with defective pituitary glands, than is a group of 2. It also depends on if you are trying to claim an non-significant difference as proof of no difference. Logically speaking, finding non-significance can never be used as evidence that two groups are the same or that a treatment had no effect (this is another gripe I have). However, people will be more willing to accept that two groups are the same, or that the treatment had no effect, if the n=200. If you use n=4, then finding a lack of a significant difference is not at all surprising, and noone would accept this as evidence that the treatment had no effect! Finally, it also depends on the novelty of the results. If the results agree with previous findings and are supported by the theory, then a smaller number of animals is generally accepted. However, results that conflict with previous, well-established results, or results which are contrary to well-accepted theories tend to require more animals before they are accepted. This is the conservatism inherent in science. >Secondly, it seems to me that when animals are more available to >use, are less expensive or more numerous, referees are more likely to >refuse papers which use a small sample size. When rare or expensive animals >are studied, it often seems adequate to report results from only one >animal (chimpanzees learning to use sign language might be a possible >example). With some reason. Usually results are judged in terms of the contribution they make to science. This is based on how much we already know, and how much more we could expect to know. If you have zero units of knowledge then an addition of 1, is a substantial increase. A study on three square-lipped rhinos performing stereotypies would probably be accepted, because we know so little about them, and we couldn't reasonably expect a study to contain many more. However, a study on broiler chickens that used only 6 birds might be criticized on the grounds that more could have been easily used. >If I taught my battery chicken (number 2,567) to bob and nod her >head in various ways to ask for a drink or food etc, would this be published >without my first replicating the study? It depends. If it took six months of round-the-clock training then it might be accepted. If it took five minutes, then the paper would probably be returned with a note suggesting resubmission after you spent the half-an-hour necessary to train the other 5! Jeff Rushen From: IN%"svj1@cornell.edu" 18-APR-1995 09:25:38.91 To: IN%"applied-ethology@sask.usask.ca" CC: Subj: equine stereotypies Dear All, I originally replied directly to Dr. Jensen regarding his question on weaving horses, then realized others may have an active interest in the discussion. The content of my reply to him is as follows... Dear Dr. Jensen, Dr. Debbie Marsden (from Uinversity of Edinburg, Scotland), is doing some very interesting work in the field of equine stereotypies and so called "stable vices". She was a recent guest here at Cornell University and she would be most helpful to you in preparing your statement. From her lecture, I will tell you that your assessment is pretty much correct. *Horses must have a predisposition to the trait, it is not learned or acquired from other horses! *It does not necessarily indicate an environmental deficiency, or ALL horses in the same environment would show some sort of "abnormal" behavior. *It seems to be related to movements associated in grazing, so a lot of horses stop when kept at pasture or when turnout time is increased. *It may be an adaptive mechanism. *Does not necessarily affect performance (the great racehorse Secretariat was a cribber!), and unfortunately, some horse people are set in some of their beliefs regarding these behaviors that they call "vices" (as if the horse is some drug addict... hehehe). For a more formal discussion of her findings, you may want to speak to Dr. Marsden directly. Sincerely, Soraya V. Juarbe-Diaz, DVM. Resident, Animal Behavior Clinic College of Veterinary Medicine Cornell University Ithaca, New York phone 607-253-3844 fax 607-253-3846 Email svj1@cornell.edu From: IN%"Anna.Olsson@ETHO.VH.WAU.NL" 18-APR-1995 09:51:33.49 To: IN%"applied-ethology@sask.usask.ca" CC: Subj: re:weaving in horses When discussing learning and 'contagiousness' of stereotypies, ethologists often stress that neighbouring animals are subject to the same stressful conditions, and that this may explain if more animals in the same stable develop stereotypies. I would like to point at another factor likely to facilitate the development of stereotypies in an animal housed next to a stereotyper: In fact, the stereotyping animal makes up quite an important part of the environment of its neighbour. Anyone who has performed behavioural observations on stereotypies in pigs knows that being in a stable together with 10 chain-chewing sows is a rather stressful experience. The same holds for weaving in horses, which when expressed at maximum is a very vigorous movement. A horse that is stabled adjacent to a weaver thus probably experiences a much more stressful environment than one whose neighbour spend the major share of the day sleeping. From: IN%"MAPPLEBY@srv0.bio.ed.ac.uk" "Mike Appleby" 18-APR-1995 10:51:26.83 To: IN%"Anna.Olsson@etho.vh.wau.nl" "anna olsson" CC: IN%"APPLIED-ETHOLOGY@sask.usask.ca" Subj: re:weaving in horses >I would like to point at another factor likely to > facilitate the development of stereotypies in an animal housed next to a > stereotyper: In fact, the stereotyping animal makes up quite an important > part of the environment of its neighbour. Anyone who has performed > behavioural observations on stereotypies in pigs knows that being in a stable > together with 10 chain-chewing sows is a rather stressful experience. The > same holds for weaving in horses, which when expressed at maximum is a very > vigorous movement. A horse that is stabled adjacent to a weaver thus probably > experiences a much more stressful environment than one whose neighbour spend > the major share of the day sleeping. This was the major conclusion of our paper on pigs: APPLEBY, M.C., LAWRENCE, A.B. & ILLIUS, A.W. 1989. Influence of neighbours on stereotypic behaviour of tethered sows. Applied Animal Behaviour Science 24: 137-146. Neighbours tended to correlate in amount of stereotypies rather than type, so we concluded that 'It seems likely that the behaviour of neighbours was an important cause of stress of newly tethered gilts.' Mike From: IN%"REPRO@nbc.upenn.edu" 18-APR-1995 11:16:34.47 To: IN%"applied-ethology-error@sask.usask.ca" CC: Subj: RE: Weaving in horses I have been working a bit on stereotypies in horses, along with Debbie Marsden of Edinburgh. I belive that clearly some horses are weavers and others are not, in the sense that some horses are far more stereotypy prone than others. Debbie has some data in the pipeline suggesting that there may be a strong genetic component to susceptibility to stereotypy. I believe a horse's past experience is also important, in terms of how stressful a given environment seems to be. It is true that a stereotypy, usually weaving, pacing, pawing or head tossing, can be induced in almost any horse, by just pressing them far enough. An example is the behavior of horses in pain. But across the typical range of domestic conditions from tie-stalls to complete pasture turn-out in social groups, some horses are stereotypy-prone and others remain stereotypy free. We have many exceptions to the classic wisdom that more "freedom" and social interaction can cure or even prevent stereotypies. On the contagiousness of stereotypies, it is true that a horse is often first observed performing a stereotypy in association with stabling or pasturing near another stereotypy-performing horse. I don't know of any work addressing other possible explanations. It's my impression that the stereotypy of one horse may be annoying or anxiety-producing in other horses. I have some observations from work with apomorphine, which induces locomotor behavior in horses. In horses in box stalls, they start circling the box immediately after IV injection. That is, except for stereotypy-prone horses, who generally perform their characteristic locomotor stereotypy. Untreated neighbor horses often appear excited or agitated by the treated horse. Stereotypy-free neighbors often start circling their box, no matter what the treated horse is doing. Stereo- typy prone untreated neighbors often begin an episode of their characteristic stereotypy. When I have had a barn full of stereotypy prone animals, it is clear that if you provoke an episode in one animal, it often sets off the whole barn, each doing their thing. So while I don't thing we fully understand this contagiousness, I certainly understand the anectdotes and conclusions of stable folks who don't want weavers and pacers in the barn. Hope this helps. Please keep us informed on the outcome. Sue M.Donnell New Bolton Center Univ PA Sch Vet Med 382 W. Street Road Kennett Square, PA 19348 610-444-5800 repro@NBC.UPENN.edu From: IN%"STOOKEY@sask.usask.ca" 18-APR-1995 11:32:32.93 To: IN%"applied-ethology@sask.usask.ca" CC: Subj: re:weaving in horses >> I would like to point at another factor likely to >> facilitate the development of stereotypies in an animal housed next to a >> stereotyper: In fact, the stereotyping animal makes up quite an important >> part of the environment of its neighbour. Anyone who has performed >> behavioural observations on stereotypies in pigs knows that being in a stable >> together with 10 chain-chewing sows is a rather stressful experience. The >> same holds for weaving in horses, which when expressed at maximum is a very >> vigorous movement. A horse that is stabled adjacent to a weaver thus probably >> experiences a much more stressful environment than one whose neighbour spend >> the major share of the day sleeping. >This was the major conclusion of our paper on pigs: >APPLEBY, M.C., LAWRENCE, A.B. & ILLIUS, A.W. 1989. Influence of neighbours on >stereotypic behaviour of tethered sows. Applied Animal Behaviour Science 24: >137-146. >Neighbours tended to correlate in amount of stereotypies rather than >type, so we concluded that 'It seems likely that the behaviour of >neighbours was an important cause of stress of newly tethered gilts.' >Mike One of the comments that I have heard a horse trainer say was that horses tend to increase the intensity, or at least start to engage in their particular stereotypic behaviour, whenever the horse's "social friend" or neighbor is removed. If true, it would mean that whatever stressful environmental factors are present by being stabled, it may also be stressful to have the neigbhor removed. I have often wondered if stabled horses would not be better off if the stable doors/windows would be adjacent to each other thereby allowing social and physical contact instead of the usual arrangement of stable doors equally spaced and all possible social and physical contact shut off. I could more easily understand, that being adjacent to another horse and at the same time being denied social and physical contact would be more stressful to a horse than the mere sight and sounds (and smells?) of another horse engaged in stereotypic behaviour. Anyone aware of a study showing a reduction of stereotypies in stabled horses that are allowed physical contact with the neighboring horse(s)? Joe =================================== Joseph M. Stookey Department of Herd Medicine and Theriogenology Western College of Veterinary Medicine University of Saskatchewan Saskatoon, Saskatchewan S7K 4Y3 Canada stookey@sask.usask.ca From: IN%"svj1@cornell.edu" 18-APR-1995 12:38:28.83 To: IN%"applied-ethology@sask.usask.ca" CC: Subj: more weaving Dear All, Joseph Stookey writes, "I have often wondered if stabled horses would not be better off if the stable doors/windows would be adjacent to each other thereby allowing social and physical contact instead of the usual arrangement of stable doors equally spaced and all possible social and physical contact shut off. I could more easily understand, that being adjacent to another horse and at the same time being denied social and physical contact would be more stressful to a horse than the mere sight and sounds (and smells?) of another horse engaged in stereotypic behaviour. Anyone aware of a study showing a reduction of stereotypies in stabled horses that are allowed physical contact with the neighboring horse(s)?" I am not aware of any studies per se that focuses on housing as it influences expression of stereotypies. Again, I would refer those interested in equine stereotypies to contact Dr. Marsden at the University of Edinburg. Her recommendation includes as much turnout as possible with a compatible herdmate. From personal experience I will say that my horse is currently housed in a tie stall arrangement where he is separated from one neighbor by a solid wall but separated from his other neighbor by a swing bar. This means that he interacts with his one neighbor quite well. He is also turned out for about 10-12hrs a day depending on weather. The schoolhorses, which do not get as much turnout, are matched in pairs by personality, and it is not unusual to see them engaged in halter-tug-of-wars, hay sharing and even a limited amount of mutual grooming. They can also lie down easily, even in lateral recumbency. The only weavers that I can recall from this barn (which houses about 30 polo horses, 20-22 school horses annd 4 instructor horses) were 2 polo horses (mother and daughter) and a warmblood in a boxstall, who gets 10-12 hrs. turnout. This warmblood's weaving intensifies at feeding time, or if she sees you with carrots or apples, this latter circumstance being anything but "stressful". She also does it in her paddock right before she is brought in. It seems an anticipatory reaction. At any rate, I have been involved with this barn since 1980, and it has never experienced an epidemic of learned stereotypies. That includes cribbers, too, of which we have had a few also. Feeding time elicits some head twisting and tongue lolling in the majority of horses, but can these activities also be called stereotypies? Soraya V. Juarbe-Diaz From: IN%"CROWELL-DAVIS.S@calc.vet.uga.edu" "Sharon Crowell-Davis" To: IN%"applied-ethology@sask.usask.ca" CC: Subj: stereotypies Regarding Per's topic of stereotypies in horses, which is certainly turning out to be popular... Having spent several thousand hours over the past 17 years trailing after domestic horses at pasture and feral horses in scrub and sitting in the corner of stalls, recording various facets of the horses' behavior, my opinion agrees with Pers', that "A stereotypy develops as a result of an interaction between an animal and its environment. An individual horse may be more or less inclinded to develop the behaviour." Observing domestic horses living in herds at pasture and feral horses on Georgia's Cumberland Island, I have been impressed with how busy they are. Subtle social interactions, invisible to the casual observer but obvious to those who know the horses well, are almost constant. Likewise, the locating of, taste-testing and sometimes continuing to consume patches of food is time-consuming. The day is constantly interrupted by major events, e.g....mutual-grooming, conflict over food patches, play, shelter-seeking, migrations from one section of a pasture to another, etc. I never observed stereotypies in horses living long-term in an open environment as members of an established social group, although some horses may develop a problem when living in a stable, then maintain the behavior when put out on pasture. I will have to disagree with the statement of Dr. Marsden, as relayed by Dr. Juarbe-Diaz that "It does not necessarily indicate an environmental deficiency or ALL horses in the same environment would show some sort of "abnormal" behavior", for a number of reasons. 1) This statement assumes that all horses in the same environment are being treated exactly the same, e.g. they are in the same size stall with the same interesting view out of the stall, are being ridden by the same person in the same way for the same amount of time at the same time of day in the same weather, are being turned out at pasture for exactly the same amount of time, etc. Obviously, different horses at the same facility are not being treated exactly the same. 2) It requires that we assume that all horses have exactly the same tolerance for environmental deficiency before we will consider this a possibility. If a Percheron were being fed rations sufficient for a 400# pony, we would consider its environment (ration) deficient. We would not use the fact that the pony in the stall next door is in good weight as justification for saying that the Percheron cannot be suffering from a ration deficiency when it is losing weight. We would likewise not pause for a moment to think that the Percheron is "abnormal" or has a "problem". We would blame the stable manager. Likewise, individual horses may have different tolerances for whatever experiences or deficiency of experiences, cause stereotypies, and still be within a normal range. Indeed, unless two horses were identical twins and had been raised in exactly the same environment, I would expect them to have different thresholds of tolerance for a deficient environment. 3) The literature on feral horses is remarkably empty of reports of weaving and other stereotypies. In 100% of the feral horses, and horses living at pasture in long-term social groups which I have observed, there have been no stereotypies. I have only observed/been informed of stereotypies developing in horses kept at least part of the time in stables. Also, see See P.D. McGreevy, P.J. Cripps, N. P. French, L.E. Green and C.J. Nicol (1995) Management factors associated with stereotypic and redirected behaviour in the Thoroughbred horse. Equine Veterinary Journal 27: 86-91. The risk of horses performing abnormal behaviour increased in association with several specific management characteristics. This included box designs which minimised contact between neighbouring horses. (to answer a more recent question about allowing the horses in neighbouring stalls contact) As for this particular case, even if we were to accept it as true that there is something "abnormal" about this particular horse, if it had never weaved before the first owners sold it, they would have had no reason to know or assume that it was any more likely than the average horse to do so. Whether or not that point is relevant would obviously depend on the laws which apply to this case. Sharon Crowell-Davis crowell-davis.s@calc.vet.uga.edu From: IN%"ir10000@hermes.cam.ac.uk" "I. Rochlitz" 18-APR-1995 13:34:22.70 To: IN%"ALund@ZI.KU.DK" "Lund, Anders {ZI}" CC: IN%"applied-ethology@sask.usask.ca" "ISAE-net" Subj: RE: Coprofagous dogs Iam not a specialist in carnivore nutrition, butto my knowledge your message contains a number of errors. The reason for cats requiring a higher protein content is not so much because of domestication, but because cats have particularly high requirements for certain amino acids such as arginine and taurine, which are found primarily in meat not vegetable protein (therefore dogs can be vegetarian, but not cats). cats can tolerate higher levels of fat in their diets than dogs (but do not develop atherosclerosis) and also have different requirements for vitamins.A cat fed a dog diet may develop problems related to the relative amino acid imbalances, although I believe that many food manufacturers supplement their dog food so that, even if the food is fed exclusively to cats, problems do not arise.And what do you mean by psychic problems? if neurologic signs develop, they may be related to the amino acid (arginine) deficiency. As far as I'm aware, dogs fed on cat food are not more prone to liver problems, in fact owners of small dogs that are poor or finicky eaters (the dogs not the owners) will often feed them cat food since it is more calorie-dense, higher in fat and therefore more palatable. You are suggesting that high protein diets cause liver problems-there is no evidence for this (high protein diets may cause behavioural problems in dogs, but that is another story). Diets are often protein-restricted for animals with renal disease (and fat-restricted for animals with liver disease) although even this recommendatoin is being revised. You state that most cats are fed more food than they need, and this results in high protein faeces. Obesity in cats certainly is a problem in about 20% of cats in the US, less in the UK, but that means that 80% of cats are not getting more food than they need. And anyway, unless the food was incredibly high in protein or very imbalanced with regards to energy content, vitamins, co-factors etc, the efficiency of digestion of the protein would not besignificantly decreased; the cats won't lose protein in their diet, they'll just put on weight.I suggest you get in touch with a small animal nutritionist, for example from Hill's or the Waltham Centre for Pet Nutrition for specialist advice. Finally, I would be interested to know the source of information for the statement that dominant cats leave their faeces in high places, and that removal of the faeces may cause psychic problems. The terms of dominance and submission should be applied with caution in describing cat behaviour, and usually when cats find urine or scent marks from other cats, they over-mark them but are not repelled, or kept off territories. Irene Rochlitz, Animal Welfare Group, Univ of Cambridge, Madingley Road, Cambridge, UK. On Tue, 18 Apr 1995, Lund, Anders {ZI} wrote: > > The question was raised if for instance the protein level in cat faeces > could motivate dogs to eat it. By putting the question to a nutrionist I got > the following answer: > It seems that domestication has reduces cats' ability to digest proteins, > therefore cat food is much higher in protein content than dog food. (The > difference in protein content in canned cat and dog food is so great, that a > cat fed with only dog food would get metabolic problems (and possibly > psychic ones?) whereas a dog fed on exclusively cat food would get liver > problems). Furthermore, pet cats usually get more food than they need. These > factors make the cat faeces high in protein and could therefore motivate the > dog to eat it. > However, there are as already mentioned numerous reports on dogs eating > faeces from pups and humans (particularly children), so it coul also be a > behaviour based on keeping the home range clean from olfactory giveaways. > In the concrete situation between a domestic dog and cat there could also be > a dominance situation: removal of teritorial markers. > Finally, dominant cats would put their faeces as high as possible, as > markers, and the continuous removal of these markers may (actually) cause > psysic disturbances in the cat and cause it to become more aggressive. > > Yours > Anders Lund > > From: IN%"RUSHENJ@NCCCOT.AGR.CA" "JEFF RUSHEN" 18-APR-1995 13:55:18.09 To: IN%"APPLIED-ETHOLOGY@sask.usask.ca" CC: Subj: weaving Vecchiotti and Galanti (Livestock Production Science 1986 14:91-95) report some evidence that weaving by horses tends to occur in families and thus may have a genetic component. The authors also suggest that cribbing is unlikely to be passed on by imitation of the mare by her foals. However, cribbing is unlikely to be related to weaving, and seems more likley to reflect some feeding problems (Gillham et al. AABS 1994 41:147-153), similar to the stereotypic behaviour of pigs. I doubt that we know enough to say whether or not horses with the pre-disposing genetic makeup would develop the behaviour regardless of what environment they were brought up in. In any case, it is surely impossible to take one animal and say with any confidence why it happened to develop a stereotypy. P. D. McGreevy at the Dept. of Animal Health, School of Veterinary Science, University of Bristolm, Langford UK (fax 0934-853443) and Christine Nicol did a very large epidemiological survey and presented the results at the ISAE meeting at Foulum. Factors which increased the likelihood of stereotypies were 1. a small amount of forage 2. non-straw bedding 3. few horses in yards 4. minimal contact between neighbouring horses in stables and 4. use of hay rather than other forage types. However, they did not report on the differences between the different forms of stereotypy. Perhaps of most relevence to Per, the abstract states that training methods did not appear to be related to the performance of the behaviour. Jeff Rushen From: IN%"robin@coape.win-uk.net" "ROBIN E WALKER" 18-APR-1995 18:37:24.63 To: IN%"Applied-Ethology@sask.usask.ca" CC: Subj: Another thread? ..This weaves itself perforce into my business Lear ii 1 17 Which is the dog. Let us take tail chasing for a moment and look at three forms. 1. A fully formed OCD conducted in isolation. 2. Determined attention seeking. 3. Superstitious whirling performed in anticipation of some reward. All of these behaviours are the result of the dog's "environment" which is a eupehemism for "us". The underlying neurology seems to be the release of psychic tension by motor routines. These tend to be typical of the species. It is most interesting to see this thread revealing that the horse is weaving because it is (to some degree) distressed by its constraints in lifestyle. It has been mentioned that horses will weave in response to routine events such as feeding time. This might mirror No. 3 for the dog. Or perhaps even No. 2. I do not see a complex issue for the brain ( "more busy than the labouring spider Weaves tedious snares" 2 Henry VI. iii 1) to grasp. It is quite obvious that animals may not all enjoy the way we force them to live but equally that a modicum of thought and ingenuity can greatly improve their lot. As You Like It Act III. Scene ii - The Forest of Arden Enter Corin and Touchstone Touch. ..Hast any philosophy in thee, shepherd? Cor. No more, but that I know..that the property of rain is to wet, and fire to burn; that good pasture makes fat sheep, and that a great cause of the night is lack of the sun. Touch..Where are thy "stats"? None shall publish this! Cor.. he that hath learned no wit by nature nor art may complain of good breeding, or comes of a very dull kindred! Touch.. Tough dugs..no graphs nor stats shall set forth naught.. Cor....Shall we go draw our numbers and set on? Touch.. Aye together and the hand of time shall this brief into as huge a volume..... From: IN%"ALund@ZI.KU.DK" "Lund, Anders {ZI}" 19-APR-1995 To: IN%"applied-ethology@sask.usask.ca" "ISAE-net" CC: Subj: RE: Coprofagous dogs Dear I. Rochlitz! Thank you very much for your insightful comment to my statement. So much for relying on others. Most of my statement was based on the reply from a cat owning nutrionist. I only checked on the question because it was raised on the net earlier. The part on psychic problems for a cat eating dog food was only a silly joke, I am sorry that did not come across. Yours Anders Lund From: IN%"pdkhuis@pobox.ruu.nl" 19-APR-1995 03:14:44.86 To: IN%"applied-ethology@sask.usask.ca" CC: Subj: World Congress Alternatives 1996 Mailing list invitation 2nd World Congress on Alternatives and Animal Use in the Life Sciences October 20-24, 1996 Utrecht, The Netherlands You are invited to attend the 2nd World Congress on Alternatives and Animal Use in the Life Sciences, October, 20-24 1996, Utrecht, The Netherlands. To be included in subsequent mailings or to update your address, please fill in the the form below and return it, by mail, fax or email, to the Congress Secretariat. You will then automatically recieve further announcements and a Call for Papers. Mailing list form Surname: Title / First name: Mailing address: Country: Phone: Fax: Email: SCOPE OF THE CONGRESS Goals: * Exchange of information on the progress made towards alternatives to animal use: - Replacement - Reduction - Refinement * Assessment of the current status and use of in vitro methodologies in toxicology, pharmacology, bioassays and basic biomedical research (advantages and limitations) * Assessment of the barriers as to the introduction of alternatives (including legislative regulations) * Exchange of information on animal welfare aspects (including procedures for the allevation of pain and distress) * Exchange of information on ethical evaluation of animal experiments and requirements for education and training * Promote the dialogue between animal protection movements and the scientific community Mailing Address World Congress Alternatives 1996 FBU Congress Bureau P.O. Box 80.125 3508 TC Utrecht The Netherlands Phone: +31.30.53.5044/2728 Fax: +31.30.53.3667 Email: l.donkers@pobox.ruu.nl s and a Call for Paper k====================================================================== Department of LAB.ANIM.SCI., UTRECHT UNIVERSITY, the Netherlands ;-} E-mail: pdkhuis@pobox.ruu.nl, tel.:*31-30-532033, fax: *31-30-537997. ***the 3 R's: REPLACEMENT, REDUCTION AND REFINEMENT OF LAB.ANIMALS*** ====================================================================== From: IN%"mjk1000@cus.cam.ac.uk" 19-APR-1995 03:37:08.07 To: IN%"applied-ethology@sask.usask.ca" CC: Subj: RE: Weaving in horses Dear All, A further two pence worth! In addition to the very valid points made by Sharon Crowell-Davis regarding objections to Dr Marsden's statement "(stereotypy performance) does not necessarily indicate an environmental deficiency or ALL horses in the same environment would show some sort of "abnormal" behavior", it is possible that two horses experiencing similar or even identical environments may use different strategies in order to attempt to cope with it. One individual may develop a stereotypy in response to adverse environmental conditions. An adjacent horse may use an alternative strategy to cope with these adverse conditions, such as becoming unresponsive to environmental stimuli and withdrawn (this has been reported in stall housed sows). Although withdrawal and unresponsiveness may not be normal behavour, it may not be as easily recognisable as abnormal as the performance of a stereotypy. Thus all horses in the same environment may indeed be showing abnormal behaviour in an attempt to cope with adverse environmental conditions, but only the stereotypy is obvious. The stereotyping horse is therefore experiencing environmental deficiences, as is the withdrawn and unresponsive horse, and both may have poor welfare. Let the discussion continue! Mark J. Kennedy, University of Cambridge, Dept. of Clinical Veterinary Medicine, Madingley Road, Cambridge CB3 OES, U.K. (mjk1000@cus.cam.ac.uk) From: IN%"ISAE.MEMBERSHIP@bbsrc.ac.uk" "ISAEMSEC" 19-APR-1995 05:19:12.05 To: IN%"applied-ethology@sask.usask.ca" "Applied Ethology" CC: Subj: ISAE Membership Info ____________________________________________________________ _ _ _ / /_ /_ / /_ / _ / / / /_ International Society for Applied Ethology ____________________________________________________________ The Society was formed in 1966 as the Society for Veterinary Ethology. It rapidly expanded to cover all applied aspects of Ethology and other Behavioural Sciences which are relevant to many human-animal interactions such as farming, wildlife management, the keeping of pets and the control of pests. It has also become increasingly international: it now has a federal structure with Regional Secretaries in several European countries, USA, Canada and Australasia. In 1991 the society was renamed the International Society for Applied Ethology. REDUCED JOURNAL SUBSCRIPTION & NEWSLETTER In a special arrangement with Elsevier Science Publishers BV ISAE Members can subscribe to the journal APPLIED ANIMAL BEHAVIOUR SCIENCE at a reduced subscription rate of 95.00 UK Pounds per year. This is approximetely one-sixth of the standard subscription charged to libraries. Members also receive a Newsletter twice a year. MEMBERSHIP Membership is open to people involved in the fields of agriculture, veterinary science, zoology and other animal sciences. The current annual membership subscription is 10.00 UK Pounds. This can be paid by personal cheque, standing order (from UK bank accounts only), bankers order, bank transfer or by Visa, Mastercard or Eurocard credit card. Further details can be obtained from the Membership Secreatary or will be sent out on confirmation of membership. A Membership Application form can be obtained from: Dr S. Mark Rutter, ISAE Membership Secretary Institute of Grassland and Environmental Research North Wyke, Okehampton, Devon EX20 2SB UK International Fax : +44 1837 82139 UK Fax : 01837 82139 e-mail : isae.membership@bbsrc.ac.uk From: IN%"Per.Jensen@hhyg.slu.se" 19-APR-1995 06:13:22.11 To: IN%"applied-ethology@sask.usask.ca" CC: Subj: RE: Functions and causes in Coprofagous dogs Dear all, Although not writing on the subject, Anders Lund nicely illustrated my concerns regarding mix of explanatory levels. On discussing why some dogs eat cat faeces, he wrote: >problems). Furthermore, pet cats usually get more food than they need. These >factors make the cat faeces high in protein and could therefore motivate the >dog to eat it. >However, there are as already mentioned numerous reports on dogs eating >faeces from pups and humans (particularly children), so it coul also be a >behaviour based on keeping the home range clean from olfactory giveaways. Clearly, these are explanations on different levels. The shit-eating may be caused by the protein contents and tastyness (jummy, jummy) of the leavings, but the function may be to keep homeranges free of olfactory cues. See what I mean? >In the concrete situation between a domestic dog and cat there could also be >a dominance situation: removal of teritorial markers. >Finally, dominant cats would put their faeces as high as possible, as >markers, and the continuous removal of these markers may (actually) cause >psysic disturbances in the cat and cause it to become more aggressive. Both of these appear to me to be functional hypotheses. Beware!! Per ******************************************************************* Per Jensen Professor of Ethology __/\______________9 Swedish University of Agricultural Sciences,\ o I Department of Animal Hygiene, \- I Section of Ethology \_______________I SKARA, SWEDEN /\ /\ E-mail: Per.Jensen@hhyg.slu.se / \ / \ ******************************************************************* From: IN%"GREENBER%TWSUVM.BITNET@cmsa.Berkeley.EDU" 19-APR-1995 06:34:06.97 To: IN%"applied-ethology@sask.usask.ca" CC: Subj: I'd like to call your attention to a journal that would be pleased to receive your manuscripts: The International Journal of Comparative Psychology Robert Hughes, Editor Psychology Department University of Canterbury Christchurch, New Zealand tel. 64 3 3642879 fax. 64 3 3642181 psych275@csc.canterbury.ac.nz From: IN%"RUSHENJ@NCCCOT.AGR.CA" "JEFF RUSHEN" 19-APR-1995 07:06:39.00 To: IN%"APPLIED-ETHOLOGY@sask.usask.ca" CC: Subj: stereotyped stereotypies ...and all the old stories come trotting out...... >One individual may develop a stereotypy in >response to adverse environmental conditions. An adjacent horse may use an >alternative strategy to cope with these adverse conditions, such as >becoming unresponsive to environmental stimuli and withdrawn (this has been >reported in stall housed sows). >Mark J. Kennedy, I am glad he used the word "reported" and not "shown". For non-English speakers, the difference is that any old thing can be "reported", whereas generally something has to be true before it can be "shown". Who says that research in stereotypies just goes round in circles?! your friendly (believe it or not) network skeptic (or should that be cynic?) Jeff Rushen From: IN%"RUSHENJ@NCCCOT.AGR.CA" "JEFF RUSHEN" 19-APR-1995 07:41:06.13 To: IN%"APPLIED-ETHOLOGY@sask.usask.ca" CC: Subj: causal coprophagy functions >Clearly, these are explanations on different levels. The shit-eating may be >caused by the protein contents and tastyness (jummy, jummy) of the >leavings, but the function may be to keep homeranges free of olfactory >cues. >See what I mean? >Per Jensen I might or I might not. He might be using the usual oversimplified way of referring to a functional hypothesis (i.e. The genetic combination that favours shit eating has spread through the population by natural selection because a consequence of this is that territories are kept clean, which is associated with higher reproductive success), or he might be talking about a causal hypothesis (i.e. Dogs have a cognitive representation of a clean environment and when their perception of their environment differs from that representation their behaviour is directed towards the goal of removing or reducing that discrepancy). Per is correct that people mix up causal and functional hypotheses. However, another problem is that people do not describe their hypothesis clearly enough, so it is hard to tell whether it is a functional or a causal hypothesis. See what I mean? Jeff Rushen From: IN%"menke@ezinfo.vmsmail.ethz.ch" 19-APR-1995 08:29:36.93 To: IN%"applied-ethology@sask.usask.ca" CC: Subj: Re leaning nose pressing dairy cows ae9501 Hello, I am a new member of the e-mail network applied-ethology and I like to give a comment to the emails from J.ROBERTS and H.HOPSTER from Jan. 21th and 26th. I studied agriculture in Germany and since 1991 I carried out an investigation about the social behaviour of horned dairy cows in loose housing systems at the ETH-Zurich. From 35 dairy herds (more than 1200 cows) I observed the social behaviour (more than 500 h.) and at no time I saw the leaning behaviour, defined in the C.E.C-report. In dehorned herds I saw this behaviour frequent, especially in the lying box or in the collecting yard where the space per cow is very low. KROHN (Proc.Soc.Vet.Ethol., 17.19.5. 1990) investigated the leaning behaviour of a dehorned dairy herd in a deep bedding barn with free access to pasture. 33% of the cows in the loose housing showed the leaning behaviour against other cows. His hypothesis is that leaning against other animals could be a type of social contact which just like social licking gives the active cow a feeling of security or expression of appeasement. I dont think so. It is known that the frequent of butting in dehorned cows is much higher than in horned cows (GRAF, 1974). The reason for this is the lower effect of a butt with the head than with horns. Often I saw that dehorned cows have no chance to push away a cow lying or standing in a lying box. In contrast to this horned dairy cows often only have to threat for pushing away the box cow. I think that leaning behaviour is a result of an unsuccessfull push away behaviour of dehorned dairy cows. Christoph Menke From: IN%"SEBM@lab0.vet.ed.ac.uk" "Sebastian McBride & Co." 19-APR-1995 08:54:48.83 To: IN%"Applied-ethology@sask.usask.ca" CC: Subj: weaving Dear all, I have read with great interest the responses to Per Jensen's enquiry about the weaving horse. We have a particular concern with strereotyped behaviour in horses here at Edinburgh, and reference has already been made to the work of Dr Deborah Marsden, so perhaps I could add a few points. From recent work that we have carried out looking at stereotyped behaviour in Przewalskis horses' there is strong evidence to support a genetic susceptability to performing stereotyped behaviour. In this regard weaving, aswell as being a behaviour that can produce problematic secondary effects ligament strain etc., it can also be considered a deficiency in the horse. Given the situation described by Per Jensen I would argue that if the horse had been knowingly sold as a 'weaver' then there is a strong case for it's return. This I would have thought is fairly difficult to prove. As already mentioned by Mark Kennedy support for stereotyped behaviour being contagious is currently still at an anecdotal level, and evidence of genetic susceptibility contradicts this theory further. However one must also say that if a horse, genetically predisposed to performing stereotyped behaviour, is placed beside another individual already crib-biting or weaving, then that horse may be promted to start the behaviour at an eariler stage because, as pointed out by Anna Olsen, this may constitute a more stressful environment. Another point that may be of interest ( particularly to Sharon Crowell-Davis) concerns some of our research carried out at the Household Cavalry Mounted Regiment in the Hyde Park barracks in London. The barracks contains approximately 180 horses all of which are Irish Draught X Thoroughbred. Nearly all the horses are kept in loose stalls, all are exercised at the same time for the same amount of time by the same people and all of them are fed at the same time with a similiar feed stuff. One couldn't ask for better experimental control (as I'm sure Paul McGreevy from Bristol could also confirm). Out of the 180 horses only 3 of them were weavers (1.7%). In the light of this I would be inclined to say that an environment may not be deficient to ALL horses, but is to some because of their, to reitterate, genetic disposition. I should also say that the rest of the horses in the barracks, apart from 3 found to crib-bite, did not appear to be coping in other fashions i.e. their behaviour appeared relativley normal. I hope these points have been of use. Sebastian McBride Dept. Vet. Clinical Studies Edinburgh From: IN%"RUSHENJ@NCCCOT.AGR.CA" "JEFF RUSHEN" 19-APR-1995 09:45:04.88 To: IN%"APPLIED-ETHOLOGY@sask.usask.ca" CC: Subj: Per is right! (re: coprophagy) >Clearly, these are explanations on different levels. The shit-eating may be >caused by the protein contents and tastyness (jummy, jummy) of the >leavings, but the function may be to keep homeranges free of olfactory >cues. >See what I mean? >Per Jensen Sorry for taking up more band-width on this, but I think Per's main point (if re-phrased) is very important and often overlooked: a causal explanation of a behaviour shouldn't depend on the consequences of the behaviour since these didn't exist when the behaviour started. It should be noted that the explanation of coprophagy as being due to protein content and the tastiness of the shit makes this mistake. For example, this cannot explain why a dog begins to eat shit since the taste is not apparent to the dog until it has begun to eat. Had Anders said "the dog eats IN ORDER TO taste the protein etc." then this would have been a causal explanation. Once the dog has begun to eat then the resulting taste may provoke it to continue eating. The protein content would most likely serve to prolong eating through the taste. An increased protein intake is a consequence of the behaviour, and so this is a functional explanation. So finally I must admit Per is probably right. Anders has mixed up a causal explanation (i.e. the dog eating the shit with the goal of keeping its territory odour-free) with a funtional one (an increased protein intake and the dog experiencing the taste as a consequence of starting to eat). Jeff Rushen From: IN%"SEBM@lab0.vet.ed.ac.uk" "Sebastian McBride & Co." 19-APR-1995 10:06:15.84 To: IN%"Applied-Ethology@sask.usask.ca" CC: Subj: Weaver Dear all, I have read with great interest the responses to Per Jensen's enquiry about the weaving horse. We have a particular concern with strereotyped behaviour in horses here at Edinburgh, and reference has already been made to the work of Dr Deborah Marsden, so perhaps I could add a few points. From recent work that we have carried out looking at stereotyped behaviour in Przewalskis horses' there is strong evidence to support a genetic susceptability to performing stereotyped behaviour. In this regard weaving, aswell as being a behaviour that can produce problematic secondary effects ligament strain etc., it can also be considered a deficiency in the horse. Given the situation described by Per Jensen I would argue that if the horse had been knowingly sold as a 'weaver' then this is unlawful and the horse should be returned. The difficulty is in proving that the horse was weaving prior to being sold. As already mentioned by Mark Kennedy support for stereotyped behaviour being contagious is currently still at an anecdotal level, and evidence of genetic susceptibility contradicts this theory further. However one must also say that if a horse, genetically predisposed to performing stereotyped behaviour, is placed beside another individual already crib-biting or weaving, then that horse may be promted to start the behaviour at an eariler stage because, as pointed out by Anna Olsen, this may constitute a more stressful environment. Another point that may be of interest ( particularly to Sharon Crowell-Davis) concerns some of our research carried out at the Household Cavalry Mounted Regiment in the Hyde Park barracks in London. The barracks contains approximately 180 horses all of which are Irish Draught X Thoroughbred. Nearly all the horses are kept in loose stalls, all are exercised at the same time for the same amount of time by the same people and all of them are fed at the same time with a similiar feed stuff. One couldn't ask for better experimental control (as I'm sure Paul McGreevy from Bristol could also confirm). Out of the 180 horses only 3 of them were weavers (1.7%). In the light of this I would be inclined to say that an environment may not be deficient to ALL horses, but is to some because of their, to reitterate, genetic disposition. I should also say that the rest of the horses in the barracks, apart from 3 found to crib-bite, did not appear to be coping in other fashions i.e. their behaviour appeared relativley normal. I hope these points have been of use. Sebastian McBride Dept. Vet. Clinical Studies Edinburgh From: IN%"CROWELL-DAVIS.S@calc.vet.uga.edu" "Sharon Crowell-Davis" To: IN%"applied-ethology@sask.usask.ca" CC: Subj: weaving In response to Sebastian McBride's letter, I remain unconvinced that weaving should be considered a "deficiency in the horse", anymore than the hypothetical Percheron of my last letter should be considered "deficient" when fed a ration suitable for a 400# animal. Due to, yes, genetics, it requires more food than a 400# animal, but the fact that it is emaciated is due to improper management, not to some deficiency of the horse. There is, of course, individual genetic variation in how much "deficiency" of environment can be experienced before an individual exhibits a stereotypy, just as there is individual variation in how much food is required to maintain weight. Indeed, genetics may affect whether or not stereotypies are exhibited at all. Some horses might endure the most horrible of circumstances-kept in tie stalls 24 hours a day with little light, poor ventilation, fouled bedding and poor food, and never exhibit stereotypies. None of this changes the fact that horses are highly social animals which, when allowed to live in a natural environment, live in relatively stable social groups of about 5-20 animals and spend a large amount of time foraging, moving about over large home ranges, and engaging in social behavior. (See Keiper, R.L. Social Structure. In (Crowell-Davis, S.L. and Houpt, K.A. (eds.) The Veterinary Clinics of North America: Equine Practice, Behavior. 2(3), 1986. for a summary of some of the data on this.) This is what numerous wild relatives of horses do as well. Being kept in social isolation in a stall, even a large stall, and exercized about an hour a day, is in no way natural for the horse. The Household Cavalry Mounted Regiment in the Hyde Park barracks in London is interesting. From the logistical viewpoint of trying to find horses of similar breeding that are maintained in as similar a fashion as possible, it sounds excellent. However, I doubt that it is possible that these horses are actually maintained in a truly identical fashion . For example: 1)"all are exercised at the same time for the same amount of time by the same people" . There are numerous possibilities for variation here. Are all 180 horses let out on the same pasture at once, at the exact same moment? (Is there a pasture that big in London?) If so, are there low-ranking horses that receive a lot of aggression from the other horses? Is there one rider who rides all 180 horses at once? (Obviously not possible.) Are there 180 riders who rotate so that they each ride the same horse for the same amount of time? Is the exact same tack used on all horses? 2) "Nearly all the horses are kept in loose stalls." Obviously not all are. Of those that are, are the stalls the exact same size? Do they have the exact same size window at the same height from which to view the outside world? Is the view of the outside world the same for all of them? Is the ventilation identical in all stalls? Are the horses unable to see or interact with each other when in their stalls? If they can see or interact with each other, we may have subordinate horses unable to avoid threats from a neighboring dominant. Do some have more access than others to interaction over the stall walls? 3) Do all of the grooms, exercizers, etc., treat all of the horses EXACTLY the same? No one has a favorite that gets an extra carrot, a couple of more minutes with the curry comb? I will stop here. Obviously, I could go on for many pages and don't really want answers to the above questions (although I am frankly curious about how one simultaneously exercizes 180 horses in London. I'd appreciate hearing about that.) The point is that even though these horses live in a similar environment, it is not possible that the environment is entirely and completely identical. We do not know whether the 3 weavers simply have a lower (genetic) threshold for living in this kind of environment or whether their environment is simply slightly more deficient than the others. (They have slightly poorer ventilation in their stalls, the grooms brush them a little less, they have slightly less turnout time, etc.) To state that since the environment is similar any differences must be genetic is to assume that small differences are irrelevant. This may not be the case. Small differences, or the summation of small differences may be critical. Even if it is true that the 3 weavers do so because of a lower (genetic) threshold of tolerance for this kind of environment, it does not mean that the horses are "deficient". I would still rest the blame entirely on the environment. (I use the term "blame" as being relevant both from a legal perspective and from the perspective of asking ourselves what is the appropriate thing to do about weavers.) Again, we have 0% weavers among feral horses which are living as a long evoluntionary process has selected them to live as opposed to 1.7% among this study group (and 3.3% abnormal behavior if add in the cribbers). Perhaps we would do better to consider the first horses in a stable that weave to be analogous to canaries in the mines. They may be individuals which are slightly more sensitive to unwholesome or deficient living conditions than others, or they may be experiencing the worst case scenario that exists in that barn, or some combination of the two. Either way, they are a sign to the managers that management practices need to be changed. As for the behaviour of the other horses in the barracks, what is "normal"? Is an absence of overt stereotypies all that is required to be "normal"? Or is it something else? While I have not collected objective data, my general impression is that horses living alone in stalls most of the time do not act the same as horses which are living on pasture in long-term social groups. Some of them, at least, do not seem normal to me. Sharon L. Crowell-Davis crowell-davis.s@calc.vet.uga.edu From: IN%"STOOKEY@sask.usask.ca" 19-APR-1995 19:09:02.46 To: IN%"applied-ethology@sask.usask.ca" CC: Subj: Weaving and welfare? Sharon Crowell-Davis wrote: (much deleted) >Perhaps we would do better to consider the first horses in a stable >that weave to be analogous to canaries in the mines. They may be >individuals which are slightly more sensitive to unwholesome or >deficient living conditions than others, or they may be experiencing >the worst case scenario that exists in that barn, or some combination >of the two. Either way, they are a sign to the managers that management >practices need to be changed. Would the first 'canary' (horse) to weave be an indication that the conditions in the 'mine' (stable) "need to be changed" for the entire group of horses in the stable or just for the horses weaving? If the expression of the stereotypic behaviour is an indicator of poor welfare, I assume the welfare of the weaving horse may be impaired, but are all the other horses in jeopardy as well? I ask this question, because I do believe that genetics plays a role (not the entire role, but is a factor) in the expression of abnormal behaviours for many of our domestic species. Therefore, I can envision that some management systems (where some of the animals show abnormal behaviour) could appear adequate, if the switch was made to a more 'tolerant' genetic stock. For example, some strains of chickens may show more feather pecking than other strains of birds show in an aviary system. We could consider the feather pecking strains to be sentinels, indicating the aviary system needs changing. Alternatively, we could switch to a strain that does not exhibit feather pecking in an aviary system. Following the canary analogy (in the strictest sense, perhaps not the way Sharon intended) suggests that an avairy system "needs to be changed" if ANY bird from ANY strain shows feather pecking if kept inside it. Following this line of reasoning seems to overlook many of the positive characteristics that have been gained through domestication, both from a human and animal perspective. I think it would be extremely difficult to come up with an adequate poultry housing system for Red Jungle Fowls if they were kept at the stocking densities that our modern poultry industry is using, but I do believe we could find genetic strains that are more 'tolerant' of the systems being used. Perhaps we are only selecting for strains that overtly exhibit tolerance, but internally remain stressed? I know this has deviated (considerably) from the original discussion on weaving, but it circles back to 'welfare' and whether one animal's inability to cope with its environment can be a sentinel for another animal's welfare? If only 3 horses out of 180 weave, it tells me that at least those 3 have a problem with the environment, but unless the other 177 are showing some other signs of their inability to cope I would say they are alright. If in fact we argue that the 3 weavers are sentinels for the other horses and the 3 are indicators that things are wrong, could we not also argue that the 177 'normal' horses are sentinels for things are right? Welfare must act at the individual level. The 3 weavers do have a problem and the 177 may or may not. However, I must admit that I doubt the other 177 horses are perfectly 'normal' in the horse-stable environment. I suspect that many of the other 177 horses have coping strategies (abnormal behaviours?) that are less noticeable or go nameless. Joe Stookey From: IN%"robin@coape.win-uk.net" "ROBIN E WALKER" 20-APR-1995 03:22:21.42 To: IN%"Applied-Ethology@sask.usask.ca" CC: Subj: weaving and Cavalry horses We seem to be tugging fierely at the same end of the worm in this correspondence:- Sharon Crowewll-Davis said:- "Observing domestic horses living in herds at pasture and feral horses on Georgia's Cumberland Island, I have been impressed with how busy they are. Subtle social interactions, invisible to the casual observer but obvious to those who know the horses well, are almost constant. Likewise, the locating of, taste-testing and sometimes continuing to consume patches of food is time-consuming. The day is constantly interrupted by major events, e.g....mutual-grooming, conflict over food patches, play, shelter-seeking, migrations from one section of a pasture to another,..." Just as our successful exploitation of the dog was facilitated by its social and species behaviours, which so easily integrate with our own (did early Man really have to work at dog domestication or was he unable to chase the pesky critters away?), so the horse seems to have been ideal for our military purposes. Dr. Crowell-Davis's description of the horses day resembles that the Troop horse. In a cavalry unit from Classical times until the present day the grooming, the play, the migrations hither and thither. so succinctly summarized above have been replaced by military routines. Stable calls, grooming, veterinary inspections, troop parades, sending horses to water, troops drills, squadron manouevres, practise in the riding halls, mounted arms drill... In her second post Dr. Crowell-Davis mentions:- "...that none of this changes the fact that horses are highly social animals which, when allowed to live in a natural environment, live in relatively stable social groups of about 5-20 animals and spend a large amount of time foraging, moving about over large home ranges, and engaging in social behavior." In Roman and Greek time the cavalry troop was around 30 strong in sub-units of 6. The Roman horses were stalled in threes at a minimum and six at maximum. The picket lines in the field has been the constant feature of troop management. Within the Roman fort the stables had dwarf walls and the horses could see from one end to the other. The social groupings described by Dr. Crowell-Davis have been replicated throughout the history of the horse in War. "I will stop here. Obviously, I could go on for many pages and don't really want answers to the above questions (although I am frankly curious about how one simultaneously exercizes 180 horses in London. I'd appreciate hearing about that.)" Dr. C-D Well, they used to wake me up for around 2 years when I worked in London. In the early morning a cavalcade would pass along the street under my window. I used to get up and count them (it's a historical OCD I have). Two troops of around 40 each went out to the Park by that route. the remaining troops will have had ceremonial Guard duties or whatever. The contingent of 180 probably represents a squadron of 4 troops and its Headquarters. You have four social groups with a tight affiliation devided into squads with are even more intimate. On manoeurvres and ceremonial drills they trot, gallop, change direction, charge, run away and lead lives oddly like they would without our interference. As to the "genetics" of fear related behaviours I am sure that Reactive and Non-reactive strains of horse could be bred just as they have been in the rat and the dog for experimental purposes. But I am unaware of any selection procedure for ceremonial cavalry horses that favours fearful behaviour, bearing in mind the nature of their duties. Similarly I am not aware that I assist my Police Dog Section to acquire "fearful" dogs. The sampling here is surely rather skewed away from individuals who will "react" to fear or frustration? Individual dog handlers and cavalrymen will always bring pressure to bear on their charges which from time to time will induce symptoms in the animal. In police dog work over 23 years I have seem stereotypies very infrequently in the German shepherd and then only in dogs being trained by rather intense or perhaps over anxious handlers. I believe that reactive behaviour in horses can be co-related to types of rider. An expert once related to an audience which included myself that childless women in their "forties" could be a problem and army officers in the "Prussian mould" could evoke reactive behaviour in dressage horses. I only mention this in passing and disclaim all responsibility for everything..it is of course a dangerous path filled with greap titfalls. Behaviour analysis must I think start with a general evaluation of the "species typicalities" and then focus via the species "fear coping or defensive repertoires" to find out what homo sapiens has been up to on each occasion. Robin Walker. From: IN%"MARYB@lab0.vet.ed.ac.uk" "Mary Booth" 20-APR-1995 03:38:56.50 To: IN%"Applied-ethology@sask.usask.ca" CC: Subj: weaving stress Dear All, I am enjoying the current discussion on stereotypies but would like to enquire about people's working definitions of "stress/stressful". Soraya V. J-D must have anticipated such a question by her use of quotation marks when she suggests that the feeding of treats (apples etc) was "anything but "stressful"" to a mare who weaved in apparent anticipation of such delights. Laying aside (but not dismissing) the possibility that this behaviour may also involve attention seeking, let us assume that the mare was thwarted in her desire to eat the apple NOW because the stable door restrained her, in "frustration" she weaved... Was the mare stressed? If so which of the following were stressors:- the apple, the anticipation, the stable door... Mary B. Mary E. Booth Dept. Vet. Clinical Studies, Vet. Field Station, Easter Bush, Roslin, Midlothian, Edinburgh, EH15 2BG U.K. Tel (0131) 650 8827 Fax (0131) 650 6588 e-mail MaryB@vet.lab0.ed.ac.uk From: IN%"RUSHENJ@NCCCOT.AGR.CA" "JEFF RUSHEN" 20-APR-1995 08:24:50.71 To: IN%"APPLIED-ETHOLOGY@sask.usask.ca" CC: Subj: stereotypies Dear Cybernauts, The present discussion on stereotypies in horses provides researchers on pigs with one of their rare opportunites to feel superior. It is pleasant to watch others flail around in a swamp that one has just managed to crawl out of. Seeing animals perform stereotypies evokes the use of value-laden terms like "stress", "deficiency" and "adverse conditions". These prejudge the issue of whether stereotypies really are a problem, and make it difficult to understand the causes. Working as it does at the interface of science and ethics, applied ethology must at some point encounter values. However, this point should be delayed as long as possible until we have a better idea of the actual causes and consequences of the behaviour. As Ian Duncan has been pointing out for years, you can't use a behavioural measure, such as the incidence of stereotypies, to measure welfare until you have some evidence that this is actually what is being measured. Deciding to use the words "stress", "adverse" and "deleterious" is not evidence. Reverting to the concept of stress in explaining stereotypies is particularly hazardous. Contributors should be aware that this issue has been nearly hacked to death, mainly through research on rats. This research is well-described in chapters by Ladewig et al., Dantzer, and Cabib in that excellent, value-for-money, everything-you-always-wanted-to-know book Stereotypic Animal Behaviour edited by A. Lawrence and J. Rushen (CABI 1993). In general, there is little *consistent* relation between the performance of stereotypies and physiological changes associated with stress, and there is little *consistent* evidence that "stressful" conditions increase the tendency to perform stereotypies. It used to be popular to relate stereos in pigs to stress in some way. This resulted in many of us floundering around in a sea of inconclusive studies trying to relate stereos to cortisol, heart rate or opioids. Little has been gained from that effort, as far as I can see, except to show that the explanation "stress produces stereotypies" does not suffer from being overly complex. My recommendation is "forget stress!". My biggest gripe, however, concerns the problem of defining stress ("Oh no!", they cry, "Don't bring THAT one up!"). Suffice it to say, that while terms like stress lack a clear consensual definition, it does not help to evoke them in explaining stereos. Since we can't define stress, since we can't agree how to measure it, and since many of us aren't sure it even exists, we can't just chortle the word "stress" when we are stumped for an explanation. I was surprised to see Mike Appleby lapse into this unhelpful mode of thought when trying to explain his interesting results on social facilitation of stereotypies. I would have thought an explanation in terms of some social facilitation of feeding motivation would have been more to the point. The word "abnormal" (and allusions to mental illness) is equally distracting since it implies that stereotypies are a class of behaviour separate from other behaviours and which are almost inexplicable. In contrast, attempts to relate stereotypies to the specific factors that cause them (to feeding problems in the case of most stereotypies by pigs), and explanations based on normal models of motivation, have proven to be most fruitful. Much of this work was done by various permutations of Terlouw, Lawrence and Appleby, who deserve considerable credit for following a focussed research programme which piece-by-piece has experimentally uncovered the specific and identifiable causes of the behaviour, rather than reverting to the wild, data-resistant speculations about "stress" that haunted this field of research. This work is described in the excellent chapter by Rushen et al in the previously mentioned excellent book, and Lawrence and Terlouw (Journal of Animal Science 1993 71:2815). The research showed that stereotypic behaviours by pigs are derived from normal behaviours, occur in the appropriate situations, and can be explained in terms of normal motivational systems. In regard to the claims that wild horses don't show stereotypies, this was often claimed for pigs, but it seems likely that post-feeding stereotypies may be closely related to, and probably derived from, rock and root chewing seen in outdoor pigs and even wild-boars. Some clues about the cause of stereotypies comes from knowing which behaviours they are derived from. From what little I have seen of weaving in stables, it looks like normal locomotion, except the horse can't get anywhere. The question is whether the horse is simply trying to locomote, or whether it is trying to get somewhere other than where it already is. The research on pigs showed the importance of experimental manipulation, rather than simply looking for correlations. Stereotypies are often found in tethered sows and so it was assumed that tethering was the cause. Now it seems most likely that tethering is irrelevant. Having said this I am very impressed with the epidemiological approach used by Paul McGreevey *provided* that it is used as a start to direct experimental tests, and is not seen as the finish. If the research on pig stereotypies had started with a careful epidemiological study, perhaps we would have avoided the erroneous conclusion that tethering was the problem. I throroughly agree with Sarah Crowell-Davis that we must be very detailed in describing the "environment" of an animal, and avoid blanket descriptions like "tethered" or "stabled". One clear conclusion of this work was that stereotypies do not indicate the general level of welfare of an animal but indicate the existence of one specific problem. In order to assess overall welfare you need to know what other problems the animal has! Obviously animals differ in the extent that they develop stereotypies, and this should be no surprise. Is this because they differ in their sensitivity to the factors that lead to stereotypy, or is it because they differ in the manner that they react to these factors? For pigs, my guess is that it is primarily the latter. Removing the causes of stereotypies, e.g. for pigs by feeding high fibre diets, can produce benefits. But it has been devilishly difficult to find any advantage, when one compares stereotyping and non-stereotyping pigs kept in the same environment. The main exception here is the amount of energy wasted, but this is be due to a direct consequence of performing the behaviour. However, one lesson that has been learnt is that speculation unsupported by any data leads us off into space. Happily, no-one has yet brought up "active and passive copers", perhaps because a horse is more obviously not a large rodent than is a pig. As a final note (What! Hasn't he finished yet?!), science can really only talk about average levels of welfare within groups of animals, and the average causes of stereotypies. We cannot take one animal performing a stereotypy and say what caused it to perform that behaviour or whether its welfare is better or worse than that of another individual animal. This requires a veterinary judgement. Jeff Rushen From: IN%"Per.Jensen@hhyg.slu.se" 20-APR-1995 11:19:53.45 To: IN%"applied-ethology@sask.usask.ca" CC: Subj: Causes and functions - shit Jeff Rushen wrote: "So finally I must admit Per is probably right. Anders has mixed up a causal explanation (i.e. the dog eating the shit with the goal of keeping its territory odour-free) with a funtional one (an increased protein intake and the dog experiencing the taste as a consequence of starting to eat). Jeff Rushen" I hope this is a joke! Jeff has challenged our minds by arguing that the consequences of a behaviour may be its cause. He has done so by assuming that the animal has a cognitive representation of its desires, and that these desires drive the animal. So, let's take a step back and ask: Why are we bothering? Is this just pure intellectual exercise? Of course not. Our capacity to formulate hypotheses and counter-hypotheses on the same level of eplanation is crucial to ethology as a science. Understanding these levels is necessary for all progress in our field. When Jeff argues (seriously or not) that the cognitive representations may drive the animal, he makes two serious mistakes. 1. It is impossible to test it. The evidence is up to you, Jeff: Give us one example of an ethological experiment that has had as a hypothesis that an animal behaved in the way it did because of being drived by a cognitive desire to obtain the consequence of the behaviour. 2. Even if it would be possible to test it, not very much had been explained at all. The reason is that we would want to know what made the animal have that cognitive desire, i e, what causal factors led it to want it, and what is the function of wanting it. I and Fred Toates (Hello, Fred! Join the discussion!) wrote a paper on behavioural needs in Appl Anim Behav Sci some time ago where we presented a conceptual motivational model incorporating the goal representation associated with a certain motivational state into a feedback loop within the system. The crucial thing about this part of the model was that the goal is evoked secondary to the causal effects of external and internal variables. Something like this: When prolactin rises in the blood of the sow, this will turn on the nest building motivation system. Once it has been turned on, the sow will have the possibility to represent the final nest as a goal, orienting the behaviour. As I have said earlier, the goal is likely often to be closely correlated with the functional consequence of the behaviour, but the causes are not necesarily so. The stickleback reacts to the red belly of the dummy. Aggressive motivation is evoked. The stickleback may now (or may not - who knows?) have a goal representation of a territory free from intruders, which is the functional reason for chasing the red stuff. But the behaviour is not caused by this goal, the behaviour AND the goal representation are caused by the causal factor "red bottom stimuli". Arguing with human examples is often striking, but misleading. When Jeff thinks he is leaving his office because he has a cognitive representation of coffee that drives him towards the coffee room, he is equally wrong (I know him - he has a coffee machine on each shelf in his ofice, he wont need to go away). Per ******************************************************************* Per Jensen Professor of Ethology __/\______________9 Swedish University of Agricultural Sciences,\ o I Department of Animal Hygiene, \- I Section of Ethology \_______________I SKARA, SWEDEN /\ /\ E-mail: Per.Jensen@hhyg.slu.se / \ / \ ******************************************************************* From: IN%"Per.Jensen@hhyg.slu.se" 20-APR-1995 11:24:37.40 To: IN%"applied-ethology@sask.usask.ca" CC: Subj: n Mike Forbes, supporting the ever-present Jeff Rushen, wrote: "The point Jeff makes that when you have 6 critters in a single enclosure you can't treat them as 6 independent subjects is dead right." A bit hard conclusion I think. What do we mean with independent subjects? Two samples are independent if the probability of (say) observing A in sample 1 is not affected by the probability of observing A in sample 2. When we flip a coin twice, the fact that we got a tails the first time tells us absolutely nothing about the probability of getting tails the second time. The two flips are clearly independent. Unfortunately, animals are not coins, and in ethology we are unlikely to ever meet the rigorous requirements of most statistical tests. Animals are simply never independent. The mere fact that you work with one species in your different treatments will cause some degree of dependence. The probability of observing behaviour A in pig 1 is correlated to the same probability in pig 2, simply because pigs are rather alike. The size of this probability is affected by numerous variables, which we always should try to minimize. Good examples are group effects and gentical effects. But it is not always possible for biological reasons to meet the requirements of statisticians. Here is an example. You want to study the effects of being in a group of a particular size and composition on the behaviour of an individual and to evaluate the group effects against individual traits such as weight and age of the individual. In order to do such an experiment, you would clearly have to keep the animals in a group but treat the individuals as the observation unit. This will probably cause an epileptic attack in your statistician, but I see no way around it. In your experimental design you would have to make all efforts to minimize group dependence, i e effects in animal A of animal B performing a certain behaviour. In many cases, you can, by biological reasoning, conclude that the behaviour of A is not particularly likely to be affected by B, and under such conditions, I suppose you can use the individual as unit. That is, six critters in an enclosure means n=6. Here is another one, relating to Jeff's original mail, where he assumed no unit below litter would ever be valid: To investigate the effect of a certain feed protein level on the growth of pigs, you get one litter of ten individuals and split them in two groups. You house them individually and feed one group with high protein food, the other with normal. Surely, your n=5 in each group even if it's only one litter? You may have problems publishing this, because the general validity may be low, but still, statistically n=5. What if you take away the walls between the single pens? You still make sure that each individual eats its ration. You may have some good independent evidence that the growth of A will not be affected particularly by the presence of B, provided they don't compete over the food. I would argue you still have an n=5, even if we now have all animals in one treatment from one litter in one group. Per PS Or I may be wrong, which Jeff Rushen will probably soon demonstrate. ******************************************************************* Per Jensen Professor of Ethology __/\______________9 Swedish University of Agricultural Sciences,\ o I Department of Animal Hygiene, \- I Section of Ethology \_______________I SKARA, SWEDEN /\ /\ E-mail: Per.Jensen@hhyg.slu.se / \ / \ ******************************************************************* From: IN%"wattsjon@herald.usask.ca" "Jon Watts" 20-APR-1995 11:45:25.96 To: IN%"applied-ethology@sask.usask.ca" CC: Subj: Tough dugs and greap titfalls Ingenious? Definitely..... Appropriate? no. Put a sock in it Robin. Jon Watts wattsjon@duke.usask.ca From: IN%"robin@coape.win-uk.net" "ROBIN E WALKER" 20-APR-1995 12:20:37.08 To: IN%"Applied-Ethology@sask.usask.ca" CC: Subj: Stereotypies "This resulted in many of us floundering around in a sea of inconclusive studies trying to relate stereos to cortisol, heart rate or opioids. Little has been gained from that effort, as far as I can see, except to show that the explanation "stress produces stereotypies" does not suffer from being overly complex. My recommendation is "forget stress!". Are not stereotypies rather like scar tissue? Are they not the result of something "injurious" rather than an injury in themselves? Is it not possible that a stereotypy is a means of "adjusting" neurendocrinological responses to "something". Might it not be the case that the neurophysiological evidence for "strain" is removed by the stereotypical behaviour? Might it not follow that in some cases the process of "adjustment" becomes in itself "rewarding" or "evoked by Pavlovian associative mechanisms"? Might the best chance of enlightenment be a study of the "insulted" organism just before it develops a stereotypical response for the first time! (When it may look rather like those organisms who are not going to develop stereotypies). Be sure to alight from the bus at the stop before I do! From: IN%"GREENBER%TWSUVM.BITNET@cmsa.Berkeley.EDU" 20-APR-1995 12:44:27.72 To: IN%"applied-ethology@sask.usask.ca" CC: Subj: Causes and functions There is an article on cognition in animals (including humans) in the recent issue of The American Psychologist. M. S. Blumberg & E. A. Wasserman (1995). Animal mind and the argument from design. American Psychologist, 50, 133-144. Gary Greenberg From: IN%"louzhen@herald.usask.ca" "Dr. Zhensheng Lou" 20-APR-1995 13:53:14.37 To: IN%"applied-ethology@sask.usask.ca" CC: Subj: Behaviour Sequencing Dear netters, I wonder if anybody in Applied Ethology have had an experience on behaviour sequencing? It seems to me that sequencing might be one of technical challanges facing all behaviourists, not only the applied behaviourists. However, in the applied ethology we see far little work on the topic. Whatever available to us has not been beyond the first-order, within-animal sequence, am I right? Anyway, I would appreciate if someone point out where to find relevant materials to read. Open discussions about it on the net are also very much welcomed. Zhensheng Lou From: IN%"wolfea@ucsub.Colorado.EDU" "WOLFE ANN B" 20-APR-1995 16:57:30.81 To: IN%"GREENBER%TWSUVM.BITNET@cmsa.Berkeley.EDU" CC: IN%"applied-ethology@sask.usask.ca" Subj: RE: Causes and functions stating that this is an article on cognition and non-human animals is somewhat of a misrepresentation. rather, the article is a behaviorist view of why animal cognition is neither valuable or should be considered a science. it is a step backwards towards the dark ages of sciences of the mind.On Thu, 20 Apr 1995 GREENBER%TWSUVM.BITNET@cmsa.Berkeley.EDU wrote: > There is an article on cognition in animals (including humans) in the > recent issue of The American Psychologist. > M. S. Blumberg & E. A. Wasserman (1995). Animal mind and the argument from > design. American Psychologist, 50, 133-144. > Gary Greenberg > From: IN%"CHRISTISON@admin.usask.ca" 20-APR-1995 17:41:29.55 To: IN%"Applied-ethology@sask.usask.ca" CC: Subj: Husbandry practices as trade barriers Can anyone give me examples of where animal management practices in one country are used as a trade barrier for the entry of animals or animal products to another country? I assume Switzerland bans the importation of battery eggs, but perhaps the ban is of all non-Swiss eggs, as opposed to specifically battery ones. Does Sweden allow entry of bacon from unbedded pigs? Britain does allow the entry of veal from closely-confined calves, I believe, but could degree of confinement be used as a barrier? What about dehorning or castration ages? I expect that there are several examples of non-therapeutic drugs permitted in one country which preclude the importation of meat to some other countries. It appears that such trade barriers are talked about much more than they exist in practice, but this may be my ignorance. Thank you for any and all responses. Iain Christison Department of Animal and Poultry Science, University of Saskatchewan. Email: Christison@admin.usask.ca. From: IN%"Per.Jensen@hhyg.slu.se" 21-APR-1995 00:13:55.50 To: IN%"applied-ethology@sask.usask.ca" CC: Subj: RE: Causes and functions Ann Wolfe wrote: >stating that this is an article on cognition and non-human animals is >somewhat of a misrepresentation. rather, the article is a behaviorist >view of why animal cognition is neither valuable or should be considered >a science. it is a step backwards towards the dark ages of sciences of >the mind. Is it really a step backwards? Has anything good and constructive come out of cognitive ethology? I have read a few exciting books, filled of anecdotes and natural history, but where are the data that advance our scientific understanding of animal behaviour? One of the few things that have changed due to cognitive ethology seems to me to be that people like Jeff Rushen these days think they can get away with claiming that behaviour is CAUSED by cognitive representations of goals. As far as I am concerned, I would like to see some more 'dark age', hard data, good-old-times ethology. Per ******************************************************************* Per Jensen Professor of Ethology __/\______________9 Swedish University of Agricultural Sciences,\ o I Department of Animal Hygiene, \- I Section of Ethology \_______________I SKARA, SWEDEN /\ /\ E-mail: Per.Jensen@hhyg.slu.se / \ / \ ******************************************************************* From: IN%"robin@coape.win-uk.net" "ROBIN E WALKER" 21-APR-1995 03:37:28.17 To: IN%"Applied-Ethology@sask.usask.ca" CC: Subj: Dear Jon Compared with the "eat shit(s_" and "get laid(s)" that have spattered recent correspondence on this group I consider my little boobs to be rather modest. There is no call for you to rudely inflate them! Try as one might a "typo" always slips past. We are not all such great poof readers as yourself. Why don't we talk about "stereotypic tail chasing in hand-reared kittens subsequently punished for fabric eating"? From: IN%"jon.cooper@zoology.oxford.ac.uk" "Jon Cooper" 21-APR-1995 04:07:37.84 To: IN%"applied-ethology@sask.usask.ca" CC: Subj: Stereotypic Behaviour. The Truth!! (fwd) ---------- Forwarded message ---------- Date: Fri, 21 Apr 1995 10:44:35 +0100 (BST) From: Jon Cooper To: gjm10@phoenix.cambridge.ac.uk Subject: Stereotypic Behaviour. The Truth!! ---------- Forwarded message ---------- Date: Thu, 20 Apr 1995 18:05:26 +0000 From: ROBIN E WALKER To: Applied-Ethology@sask.usask.ca Subject: Stereotypies Dear Robin and anyone else who has taken offence at Jeff "Voice of Reason" Rushen's commentry on do's and do not's of research into stereotypic behaviour, "This resulted in many of us floundering around in a sea of inconclusive studies trying to relate stereos to cortisol, heart rate or opioids. Little has been gained from that effort, as far as I can see, except to show that the explanation "stress produces stereotypies" does not suffer from being overly complex. My recommendation is "forget stress!". To suggest that relating stereotypic behaviour to stress is not the heresy it may at first glance appear. Stereotypies are simply defined as repetitive, invariant and with no purpose. As such they cover a huge range of activities, performed by lots of different animals in a wide range if situations. They are often found in animals in unnatural environments, particularly those that restrict the behavioural repertoire, so it is often assumed that they are bad and an indicator of a poor environment. This view gained more credance when it was found that opioid blockers reduced the incidence of stereotypies (along with lots of other behavioural responses) and opioids were involved in the modulation of stress. Stress is also generally thought of as a bad thing, although the neurophysiological parameteres used to measure it can be induced by both positive and negative challenges on the animal. It may be no coincidense that these two "bad" things (high physiological correlates of stress and tendancy to repeatedly do pointless things) are often found in restrictively housed animals, but until a causal mechanism is demonstrated, to suggest that stereotypies are caused by stress is just one of a number of potential explanations. Where detailed studies have been carried out (and some pretty good studies have been carried out in pigs, mice and errm... voles), the evidence of a causal relationship is far from consistant Check out the papers Jeff suggested and it will become very much clearer. Q1 Are not stereotypies rather like scar tissue? Are they not the result of something "injurious" rather than an injury in themselves? Stereotypies are a response to a restrictive environment. This does not necessarilly imply that the initial challenge that caused the stereotypy was injurous, stressful or noxious. Q2 Is it not possible that a stereotypy is a means of "adjusting" neurendocrinological responses to "something". Yes in the same way that all behaviour are a response to something and will be mediated by and have effects on neurophysiological control mecanisms. If I was a pig in an outdoor system and my food-supply was exhausted and I was still motivated to feed, I'd start walking about, rooting and chewing on available substrates like maybe stones. If I was a pig chained in a stall and my food supply was exhausted and I was still motivated to feed, I may start "vacuum" rooting on the bare concrete floor and chewing on any avilable substrates like maybe the bars. Same underlying physiological mecanism involved but channeled by the environment into semmingly more or less appropriate or pointless activities by the environment. The pigs in the stalls may show higherlevals of physiological stress measures, but as mentoined above this is insufficient to say this is related to stereotypy. Where it has been investigated there is little or no evidence that performance of bar-biting or sham rooting modulates these correlates. Might it not be the case that the neurophysiological evidence for "strain" is removed by the stereotypical behaviour? Could be. One way of investigating this is to take away the stereotypy and see what then happens to our measures of strain. Again inconsistant eveidence. No effect of preventing chain chewing in pigs, some effect (prolonged elevation of some hormaonal and white blood cell measures) in mice and voles. Could be that preventing any behaviour would have had the same effect. Could be that some activities (locomotor stereotypies in rodents) are related to mediation of certain aspects of (I wish I had a better word for this) stress, whilst others (oral stereotypies in food deprived animals) are mediated by the mecanisms that control feeding behaviour. Might it not follow that in some cases the process of "adjustment" becomes in itself "rewarding" or "evoked by Pavlovian associative mechanisms"? Again could be. It seems reasonable that some stereotypies are a learnt response to the environment, so can be explained both in terms of the rewarding nature of their consequences and one such reward could be reduction in the aversiveness of the environment.Again there are alternative explanations, which do not rely on the rewarding nature of a behaviour.A stimulus-response model of the control of behaviour can be used to explain the performance of stereotypies without the complication of positive or negative feedback or rewards until it is demonstarted otherwise. At a neurophysiological leval, it appears that simple repeated stimulation of a particualr neural pathway can increase its sensitivity and consequently the liklihood of future initiation. No need for rewards here either. Might the best chance of enlightenment be a study of the "insulted" organism just before it develops a stereotypical response for the first time! (When it may look rather like those organisms who are not going to develop stereotypies). Yes! Yes!! Yes!!! The major failing in research into stereotypies has beeen the small number of developmental studies. Only two teething problems, being confidant that you can produce stereotypic and non-stereotypic inviduals in the same environment, which isn't that tricky really, and the ethics of putting animals into an environment where you know they will develop a behaviour which may (or may not) be an indicator of poor welfare. Having wrestled with these problems. (The end justifies the means etc.) there's a lot a potential in investigating the causes oif stereotypy. Without studies like this its pretty easy to wildly speculate about the causes and effects of stereotypic behaviour, especially when most people only get to notice them once they are really obvious, by which time the causes could well be intangible. Be sure to alight from the bus at the stop before I do! Why? I'll choose my own bus stop thank you. Yours Jonathan Cooper From: IN%"RUSHENJ@NCCCOT.AGR.CA" "JEFF RUSHEN" 21-APR-1995 09:03:40.25 To: IN%"APPLIED-ETHOLOGY@sask.usask.ca" CC: Subj: cognition (NO LONGER "cause and function"!) >people like Jeff Rushen these days think they can get away >with claiming that behaviour is CAUSED by cognitive representations of >goals. >Per Ah, there's nothing like misrepresenting your opponent's views as a way of winning an arguement. I seem to have entered Per's imagination as a fantasy figure who says all of the things that clearly upset him. What I SAID was: explanations of behaviour that hypothesize purposes, or use the term "in order to", might be causal explanations because the hypothesized causal factors (the mental representation of the goal, and the desire to achieve the goal) do exist at the time that the behaviour started. They are different from teleological explanations that invoke as causal factors things that only exist as a result of the behaviour being performed. As an example of the latter: we cannot say that the taste of the (OK all you prudes!) "feces" caused the dog to start eating it, because the taste only comes as a consequence of the dog eating. The dog may have eaten "feces" before and remembers that it tastes nice. But now it is the memory of the taste that is the causal factor involved in motivating the dog to start eating. The point of my arguement was that Ian Duncan's "in order to" criterion is not useful in distinguishing causal from functional explanations, and that Per's assumption that explanations dealing with goals are necessarily teleological, is not correct. Nowhere did I claim that such explanations were useful, testable, or correct. I can't speak for people "like" me, but I believe that the validity of explanations of behaviour that invoke cognitive representations of goals must be tested on a case-by-case basis. That is, we have to look at each specific cognitive explanations of each specific behaviour, examine how well that it corresponds with the data that we have, and see if it IS possible to have a superior explanation that does not invoke cognitive processes. Claims that "we can always explain the behaviour as well as or better without cognitive factors" (e.g. McFarland 1989) are best considered as a "no you can't! Yes, I can!" dialogue. Dozens of books have been published and hundreds of papers written. We can't just "read a few books" and decide the whole field is worthless. Per's claim that we must eschew such explanations "on principle" is a simple throwback to the dark ages of behaviourism. (Interestingly such "in principle" rejections of cognitive ethology seem to come most frequently from ethologists interested in functional questions. Could it be that their obsession with the function of a behaviour has led them to neglect all the developments that have occurred in causal explanations of behaviour as a result of decades of research into neuroscience, artificial intelligence, cognition, learning, the philosophy of the mind etc.?). Per is asking us to revert to a "good old days" version of ethology, locked in the past, isolated and insulated from the intellectual currents of the "cognitive neurosciences" swirling around us. Unfortunately, too many applied ethologists have not kept up with what is now the intellectual core of the study of behaviour. Some of the philosophical changes are well explained in P. Churchland Neurophilosophy 1986. MIT Press 1986. One major advantage in proposing cognitive explanations of behaviour is that the attempts to test them lead us to discover interesting things about behaviour and to fully appreciate its complexity. This is the main arguement for starting off with a cognitive explanation. What is the problem if, after having made these interesting discoveries, we occassionally find that we can explain them in terms of some other theory? Many of the papers in the book by C. Ristau 1991 Cognitive Ethology Lawrence Erlbaum demonstrate how ethology is changing and beginning to deal with the complexity of behaviour. This approach is far more stimulating that the deadening behaviourism that Per is asking us to regress to. Jeff Rushen From: IN%"mrenner@wcupa.edu" "Renner, Michael" 21-APR-1995 09:05:53.06 To: IN%"Applied-ethology@sask.usask.ca" "'Applied Ethology List'" CC: Subj: FW: n Per Jensen wrote: " . . .in ethology we are unlikely to ever meet the rigorous requirements of most statistical tests. Animals are simply never independent. The mere fact that you work with one species in your different treatments will cause some degree of dependence. The probability of observing behaviour A in pig 1 is correlated to the same probability in pig 2, simply because pigs are rather alike. . ." (A buzzer sounds. The host says "Sorry, but thanks for playing.") This is an utterly spurious argument. The mathematics underlying the statistical tests in question couldn't care less whether there is some CONCEPTUAL or theoretical reason to believe the scores will be alike. Rather, the issue is whether the phenomenon (i.e., behavior) underlying one observation directly influences -- i.e., changes the probabilities of possible outcomes -- one or more of the other observations. If the pigs are in the same enclosure, their behaviors aren't independent by this test (unless they're all comatose). If they are housed separately most people would accept that the behavior of one pig doesn't _substantially_ affect the behaviors of the others, and so their behaviors are independent in the statistical sense. Later in his post, Per says much the same thing: " . . .In many cases, you can, by biological reasoning, conclude that the behaviour of A is not particularly likely to be affected by B, and under such conditions, I suppose you can use the individual as unit. . ." but it's not at all clear to me from the rest of the message that this reasoning is being consistently applied. It seems reasonable that you might want social species to be visually isolated and auditorily isolated from one another. If you really want to be obsessive about it, you could also insist on separate air supplies, etc., but that's where some expertise in the discipline plays a role. If there's no empirical or rational reason to expect that certain variables (e.g., olfaction) would affect the variables in which you're interested, it may be tactically sensible, cost-effective, and simply more efficient to ignore them. Otherwise, you end up with old-style parametric research, where you must try out all possible combinations of all variables, all the while controlling for everything. This is obviously a lousy approach. The same arguments would apply whether you were dealing with pigs, iron filings, oak trees, or college sophomores, but every discipline has some variant on the argument Per uses, which boils down to "Yes, but OUR area is so much more complex than the other discplines that we simply can't do things properly." Everybody uses this argument, and it's no more sensible when we do it than when "they" do. If you violate the assumptions of a statistical test, the numbers it produces may well be garbage. Any decision you make based on those numbers will, of necessity, be at least risky and quite possibly wrong. ---------------------------------------------------------------------------- ----- Michael J. Renner Department of Psychology West Chester University West Chester, PA 19383 Voice: 610-436-2925 Fax: 610-436-3150 Internet: MRenner@Wcupa.Edu ---------- From: applied-ethology-error To: applied-ethology Subject: n Date: Thursday, April 20, 1995 7:12PM Mike Forbes, supporting the ever-present Jeff Rushen, wrote: "The point Jeff makes that when you have 6 critters in a single enclosure you can't treat them as 6 independent subjects is dead right." A bit hard conclusion I think. What do we mean with independent subjects? Two samples are independent if the probability of (say) observing A in sample 1 is not affected by the probability of observing A in sample 2. When we flip a coin twice, the fact that we got a tails the first time tells us absolutely nothing about the probability of getting tails the second time. The two flips are clearly independent. Unfortunately, animals are not coins, and in ethology we are unlikely to ever meet the rigorous requirements of most statistical tests. Animals are simply never independent. The mere fact that you work with one species in your different treatments will cause some degree of dependence. The probability of observing behaviour A in pig 1 is correlated to the same probability in pig 2, simply because pigs are rather alike. The size of this probability is affected by numerous variables, which we always should try to minimize. Good examples are group effects and gentical effects. But it is not always possible for biological reasons to meet the requirements of statisticians. Here is an example. You want to study the effects of being in a group of a particular size and composition on the behaviour of an individual and to evaluate the group effects against individual traits such as weight and age of the individual. In order to do such an experiment, you would clearly have to keep the animals in a group but treat the individuals as the observation unit. This will probably cause an epileptic attack in your statistician, but I see no way around it. In your experimental design you would have to make all efforts to minimize group dependence, i e effects in animal A of animal B performing a certain behaviour. In many cases, you can, by biological reasoning, conclude that the behaviour of A is not particularly likely to be affected by B, and under such conditions, I suppose you can use the individual as unit. That is, six critters in an enclosure means n=6. Here is another one, relating to Jeff's original mail, where he assumed no unit below litter would ever be valid: To investigate the effect of a certain feed protein level on the growth of pigs, you get one litter of ten individuals and split them in two groups. You house them individually and feed one group with high protein food, the other with normal. Surely, your n=5 in each group even if it's only one litter? You may have problems publishing this, because the general validity may be low, but still, statistically n=5. What if you take away the walls between the single pens? You still make sure that each individual eats its ration. You may have some good independent evidence that the growth of A will not be affected particularly by the presence of B, provided they don't compete over the food. I would argue you still have an n=5, even if we now have all animals in one treatment from one litter in one group. Per PS Or I may be wrong, which Jeff Rushen will probably soon demonstrate. ******************************************************************* Per Jensen Professor of Ethology __/\______________9 Swedish University of Agricultural Sciences,\ o I Department of Animal Hygiene, \- I Section of Ethology \_______________I SKARA, SWEDEN /\ /\ E-mail: Per.Jensen@hhyg.slu.se / \ / \ ******************************************************************* From: IN%"mrenner@wcupa.edu" "Renner, Michael" 21-APR-1995 09:13:14.17 To: IN%"Applied-ethology@sask.usask.ca" "'Applied Ethology List'" CC: Subj: RE: Behaviour Sequencing On April 20, Zhensheng Lou wrote: "I wonder if anybody in Applied Ethology have had an experience on behaviour sequencing? . . ." -------------------------------- I've done some work on sequences in behavior. Specifically, I've been looking at the issue of act sequences in investigatory behavior directed towards objects as they change from novel to familiar. (The initial report was just published in the J. Comparative Psychology, 1994, v108(4), 335-343. It's a companion to a paper analysing the same data set at a more macro-level, published in JCP in 1991.) Sequencing is a real technical challenge, because you have to decide how to parse a behavior string into its components before you have anything to sequence. If you parse the string differently than the animal's system -- whatever it is -- did, your answer may look wildly different from its real underlying structure (or action plan, motor program, cognitive map, or whatever it may be). Before one can reasonably look for population-level sequence rules, it seems necessary to first demonstrate that the behavior of individuals has a non-random sequential nature that can be described in an orderly way. (I know we ALL believe this, but what evidence is there for it, really?) To do this, we've used a quasi-grammatical logic, looking for transition rules between behaviors. We've looked at such things as consistency within individuals over sessions, transfer of sequences from familiar objects to novel ones, and whether "grammars" derived from the behaviors of one animal can be successfully applied to the behaviors of other animals. These are tiny steps toward the goal you have in mind. Obbviously, I think you've raised important issues and identified a major gap in our understanding of behavior. I'd be happy to discuss these issue further with anyone who is interested, either privately or via group discussion. Michael Renner ---------------------------------------------------------------------------- ----- Michael J. Renner Department of Psychology West Chester University West Chester, PA 19383 Voice: 610-436-2925 Fax: 610-436-3150 Internet: MRenner@Wcupa.Edu From: IN%"RUSHENJ@NCCCOT.AGR.CA" "JEFF RUSHEN" 21-APR-1995 12:51:17.65 To: IN%"APPLIED-ETHOLOGY@sask.usask.ca" CC: Subj: n on 'n on 'n on! >Mike Forbes (Wasn't it Michael Renner?) >supporting the ever-present Jeff Rushen you're not doing too badly yourself Per! >To investigate the effect of a certain feed protein level on the growth of >pigs, you get one litter of ten individuals and split them in two groups. >You house them individually and feed one group with high protein food, the >other with normal. Surely, your n=5 in each group even if it's only one >litter? You may have problems publishing this, because the general validity >may be low, but still, statistically n=5. Statistically an n=5 is better than an n=4, in that the sample mean is a more accurate estimate of the population mean. Is it in this case? Let's do a thought experiment. Imagine there are two scientists working independently on this problem. One scientist (Lets call him "Michael") starts the experiment with his n=2. Unbeknown to him, all piglets were from one litter. Also, unbeknown to him, the sow of this litter had a gene which makes her unable to digest this particular protein source, and this is passed onto all the piglets. He does his experiment and finds no effect. However, he realizes that an n=2 is small and that there is a degree of error in estimating the population response (the response of all pigs) to that diet from his sample response. He knows that this error is smaller as his sample size increases. He decides to add another 2 pigs per group which gives a n=4, and he knows that this sample mean will give a better estimate of the population mean and that he will be more confident of the results. To increase his n, he choses 2 piglets, carefully ensuring that they are from different litters. Since the genetic malady is rare, these piglets do respond and he finds an overall response to the diet. Since n=4 gives a more accurate estimate than n=2 of the population response, he accepts the result of the n=4, concludes that, on average, pigs do respond to this diet. He submits his paper. The other scientist (Let's call him "Per") does the same experiment, and by chance also uses 2 littermates per treatment, of piglets that are genetically unable to respond to the diet. He finds no response with his n=2. Like Michael, he is also troubled by small sample sizes, but decides to go for n=5 instead. He takes 3 piglets, but is not careful about how they are selected, and ends up with 3 more piglets from the same litter, which cannot absorb the protein. He repeats the experiment and still finds no effect. He concludes that pigs in general do not respond to the diet, and submits his paper. One reviewer happens to receive both these papers, with contradictory conclusions. He knows n=5 provides a better estimate of the population response than does an n=4. He rejects Michaels paper and accepts Per's. In this case, we can forgive Per and we can can forgive the reviewer for their mistakes because they did not know that the pigs were drawn from the same litter. But imagine how I feel when I am the reviewer and the author KNOWS and ADMITS that he has chosen littermates to increase his n!! (In case you are wondering why I am wasting so much time writing these emails- I am hoping to reduce the time I waste reviewing flawed papers!) >What if you take away the walls between the single pens? You still make >sure that each individual eats its ration. You may have some good >independent evidence that the growth of A will not be affected particularly >by the presence of B, provided they don't compete over the food. I would >argue you still have an n=5, even if we now have all animals in one >treatment from one litter in one group. God may have good independent evidence that the growth of A is independent of the growth of B. But humans do experiments and they may not know that this diet increases aggression so that the larger A gets, the more he beats up B, resulting in higher cortisol which reduces the growth of B. >Or I may be wrong, which Jeff Rushen will probably soon demonstrate. But will he admit it? Jeff Rushen From: IN%"harrism@sask.usask.ca" 21-APR-1995 13:21:06.13 To: IN%"applied-ethology@sask.usask.ca" "etho-list" CC: IN%"harrism@sask.usask.ca" "u -- harrism" Subj: ... list moderation? At present, applied-ethology is an unmoderated list. Whatever is posted goes directly to the list readers - meaning minor inconveniences like misdirected "subscribe" and "unsubscribe" requests, and also more major hazards, such as sexist, racist, mysoginistic postings. On some lists, messages go first to a moderator, who among other things, may correct grammar, or reject messages if they contain obviously offensive material. What to you think, gentle readers? Would list moderation for applied-ethology be useful or not? A good step, or a step towards thought policing? - Moira Harris From: IN%"Joy_A_MENCH@umail.umd.edu" 21-APR-1995 14:12:24.72 To: IN%"applied-ethology-error@sask.usask.ca" CC: Subj: RE: Breasts To Moira Harris and the other members of the applied ethology list: Dear Moira, No, you are not alone. The tenor of some (and I emphasize only some) of the conversations on this net recently has resembled a "boys club" more than a forum for discussion of important topics in applied ethology. There was an interesting article in the Washington Post last week written by a woman who is an expert on the "information superhighway." She stated that women are defecting from the net (or failing to post on it, as Joe Stookey demonstrated in his article in the last ISAE Newsletter) in record numbers precisely because of this type of atmosphere. Certainly there is room on an applied ethology network for discussions of human ethology, and the causes and functions of various aspects of human behavior as well as physical traits. However, it is also possible to phrase these discussions in terms that are not offensive to women (or to those men on the net who have also made comments about, or been perturbed by, the recent tone). So please, cease---and let us return to polite and productive discussion. Joy Mench From: IN%"Joy_A_MENCH@umail.umd.edu" 21-APR-1995 15:26:33.19 To: IN%"applied-ethology-error@sask.usask.ca" CC: Subj: RE: Breasts To Moira Harris and the other members of the applied ethology list: Dear Moira, No, you are not alone. The tenor of some (and I emphasize only some) of the conversations on this net recently has resembled a "boys club" more than a forum for discussion of important topics in applied ethology. There was an interesting article in the Washington Post last week written by a woman who is an expert on the "information superhighway." She stated that women are defecting from the net (or failing to post on it, as Joe Stookey demonstrated in his article in the last ISAE Newsletter) in record numbers precisely because of that type of atmosphere. Certainly there is room on an applied ethology network for discussions of human ethology, and the causes and functions of various aspects of human behavior as well as physical traits. However, it is also possible to phrase these discussions in terms that are not offensive or demeaning to women (or to those men on the net who have also made comments about, or been perturbed by, the recent tone). So please, cease---and let us return to polite, sensitive, and productive discussion. Joy Mench From: IN%"mheeb@darwin.bio.uci.edu" "Michaela Heeb" 21-APR-1995 16:24:01.32 To: IN%"harrism@sask.usask.ca" CC: IN%"applied-ethology@sask.usask.ca" "etho-list", IN%"harrism@sask.usask.ca" "u -- harrism" Subj: RE: ... list moderation? No you are not the only one who is/was offended - a couple of weeks back I posted a request that R. W. curtail his use of inappropriate language. . .this was met with a fair number of personal replies that congratulated me on my posting (indicative that many individuals have been repeatedly offended) and also by a handful of personal (and one to the BB) that my being offended was somehow childish? Amazing attitudes for scientists in the 1990s??? I do think it is important to remember that at this point it appears to be a single individual and I now simply delete his mail without regard to content - this may seem inappropriate to some of you; but a couple of personal messages from R. W. were even more offensive than what has appeared on the net and I have no desire to spend any energy trying to filter out any information that may be of interest to me. I truly enjoy most of the posts to this BB and of the other science BBs that I subscribe to and I feel the use of my delete key to be far safer than someone moderating all of our messages because of a single individuals actions. To those who bashed my last message - please don't bother doing it again - it's hard enough being a scientist without having to also filter through this "stuff". Lets get back to science - the threads have been great lately!!!! Thank you Missy (a.k.a. Michaela) \o/ --/~\-- |0 0| / \ / \ mheeb@darwin.bio.uci.edu / \ Dept. of Psychobiology | | University of California \ / Irvine \ / | |________/ From: IN%"LPINHEIRO@APS.UoGuelph.CA" 21-APR-1995 16:51:37.86 To: IN%"applied-ethology@sask.usask.ca" CC: Subj: RE: ... list moderation? Dear Moira, dear all: About Moira's suggestion: > > At present, applied-ethology is an unmoderated list. Whatever is posted > goes directly to the list readers - meaning minor inconveniences like > misdirected "subscribe" and "unsubscribe" requests, and also more major > hazards, such as sexist, racist, mysoginistic postings. > > On some lists, messages go first to a moderator, who among other things, > may correct grammar, or reject messages if they contain obviously offensive > material. > > What to you think, gentle readers? Would list moderation for > applied-ethology be useful or not? A good step, or a step towards thought > policing? > > - Moira Harris I believe your suggestion is made with the best of the intentions, and I can understand how offended some or many of us are by some comments in the net. However, "moderation" for me is a kind of censorship, and I am against censorship. The best moderator I know, as far as my own values are concerned, is myself... So, I "moderate" and "filter" what I want to read. If I am feed up of a specific person's comments, I may just delete it. If I already had too many "n"s, causes and functions, I may not read that subject. As for the eventual "offenders", the worst punishment for them would be if we just IGNORE them. After all, everything they (or he, or she) want(s) is attention... Best regards, L. Carlos Pinheiro Machado Filho. Luiz Carlos Pinheiro Machado Filho Univ. Federal de Santa Catarina University of Guelph CCA - Dep. de Zootecnia Dep. of Animal & Poultry Sci. Florianopolis, S.C. - BRASIL Guelph, ONT - CANADA Bolsista CNPq - Brasil Sponsored by CNPq - Brazil From: IN%"mrenner@wcupa.edu" "Renner, Michael" 21-APR-1995 19:27:58.62 To: IN%"Applied-ethology@sask.usask.ca" "'Applied Ethology List'" CC: Subj: FW: ... list moderation? On April 21, "Missy (a.k.a. Michaela)" -- I infer the last name to be Heeb, but apoligize if I've mis-parsed the address -- wrote: "... was met ... by a handful of personal (and one to the BB) that my being offended was somehow childish?" ------ I was the one who posted the reply to the list rather than sending it to the "poster" individually, because I think it is a topic that deserves open discussion by this community. I did not wish to offend anyone, but what seems blunt and straightforward in one's mind or oral conversation may seem unkind when read from the printed word; if that's what happened, please accept that my intentions were to stake out a clear position in a debate, not belittle anyone. Missy, we agree about the core issue, and you missed the point of what I was trying to say; my fault for not making myself plain. I also chose a public reply because the poster had chosen a public forum for castigating the source of the message that provided the original "offense." It seems somehow asymmetrical to broadcast your derision of someone for posting a message, and then feign offense that replies come via public forum. I think -- my opinion only, and I'm prepared to be ignored or outvoted -- that any form of censorship, filtering, or "moderation" is not in keeping with an open academic discussion. As we do in face to face conversation, faculty meetings, and broadcast political debate, we must each learn to quit paying attention to those whose views are repugnant or incoherent from our frame of reference. I know that means you have to read some crap -- and those of use with lame mail systems like Microsoft mail can't even tell who the original poster is unless they sign it -- but that's the price of free speech. I would rather pay this price than have someone else decide what opinions and ideas I get to see, no matter how well-intentioned that person is, nor how closely their judgments about the appropriateness of a message might correspond to my own. If anyone wants to berate me about this opinion, it's your right, but please feel free to do so directly unless you are trying to contribute to forming a consensus (or majority) opinion. Michael Renner MRenner@Wcupa.Edu From: IN%"seasley@nmsu.edu" "Stephen Easley" 21-APR-1995 21:28:56.72 To: IN%"Applied-ethology@sask.usask.ca" "Applied Ethology List" CC: Subj: Three Sisters Three Sisters of Saskatoon Three sisters, Joy, Moira, and Missy As language police got quite pissy Perhaps they're delirious To pen screed so serious On matters not weighty, just prissy. Have a humor-filled day. :-) Stephen Easley New Mexico State University seasley@nmsu.edu