Tag Archives: Messerli Research Institute of the Vetmeduni Vienna

Dogs can adopt the perspective of humans

Humans are able to interpret the behaviour of others by attributing mental states to them (and to themselves). By adopting the perspectives of other persons, they can assume their emotions, needs and intentions and react accordingly.

In the animal kingdom, the ability to attribute mental states (Theory of Mind) is a highly contentious issue. Cognitive biologists from the Messerli Research Institute of the University of Veterinary Medicine Vienna could prove with a new test procedure that dogs are not only able to identify whether a human has an eye on a food source and, therefore, knows where the food has been hidden. They can also apply this knowledge in order to correctly interpret cues by humans and find food they cannot see themselves.

This perspective taking ability is an important component of social intelligence. It helps dogs to cope with the human environment. The results have been published in the journal Animal Cognition.

 

The so-called Theory of Mind describes the ability in humans to understand mental states in conspecifics such as emotions, intentions, knowledge, beliefs and desires. This ability develops in humans within the first four or five years of life while it is usually denied in animals. Indications that animals can understand mental states or even states of knowledge of others have only been found in apes and corvids so far. Dogs have been tested several times, but the results were poor and contradictory.

With a new experimental approach, cognitive biologists from the Messerli Research Institute could now provide solid evidence for dogs being able to adopt our perspective. By adopting the position of a human and following their gaze, dogs understand what the human could see and, consequently, know.

This ability to ascribe knowledge is only a component of a full-blown Theory of Mind, but an important one.

Identifying the right informant

The so-called Guesser-Knower paradigm is a standard test in research into the attribution of knowledge to others. This experiment involves two persons: a “Knower” who hides food, invisibly for the dog, in one of several food containers or knows where somebody else has hided it, and a “Guesser”. The Guesser has either not been in the room or covered her eyes during the hiding of the food. A non-transparent wall blocks the animals’ view of the food being hidden. After that, the two humans become informants by pointing to different food containers.

The Knower always points to the baited container and the Guesser to another one. All containers smell of food. “To get the food, the dogs have to understand who knows the hiding place (Knower) and who does not and can, therefore, only guess (Guesser). They must identify the informant they can rely on if they have to decide for one food container,” said principal investigator Ludwig Huber. In approximately 70 per cent of the cases the dogs chose the container indicated by the Knower – and thus were able to successfully accomplish the test. This result was independent of the position of the food container, the person acting as the Knower and where the Guesser was looking.

Eye on hidden food source

Dogs are able to identify the human having an eye on a hidden food source. (Photo: Ludwig Huber/Vetmeduni Vienna)

Dogs can adopt human perspectives

The only aim of this test series, however, was to independently confirm a study carried out in New Zealand. Clear evidence of dogs being able to adopt our perspective and take advantage of it was provided in a new test developed by the team, the so-called “Guesser looking away” test.

In this new experiment, a third person in the middle hides the food. This person does not give cues later on. The potential informants were kneeing left and right of this hider and looked to the same side and slightly down. Thus, one of the two persons looked towards the baiter, the other person looked away. “This means that the tested dogs, in order to get the food, had to judge who is the Knower by adopting the informants’ perspectives and following their gazes,” explained Huber. Even in this test, which is very difficult for the animals, approximately 70 per cent of the trials had been mastered.

Adopting the human perspective leads to invisible food

Being able to adopt the perspective of a human does, however, not require the ability to understand intentions or wishes. “But the study showed that dogs can find out what humans or conspecifics can or cannot see,” explained Huber. “By adopting the positions of humans and following their gazes geometrically, they find out what humans see and, therefore, know – and consequently whom they can trust or not.”

In similar experiments, chimpanzees and few bird species such as scrub jays and ravens were able to understand the state of knowledge and also the intentions of conspecifics and modify their own behaviour accordingly. For dogs, there have only been specualtions and vague indications so far. But dogs understand our behaviour very well, for example our degree of attention. They can learn from directly visible cues such as gestures or gazes. Thus, they are able to find food even if their view of it has been blocked. “The ability to interpret our behaviour and anticipate our intentions, which has obviously developed through a combination of domestication and individual experience, seems to have supported the ability to adopt our perspective,” said Huber. “It still remains unclear which cognitive mechanisms contribute to this ability. But it helps dogs to find their way in our world very well.”

Source:  University of Veterinary Medicine, Vienna media release

Advertisements

Dogs share food with other dogs even in complex situations

Generosity, even among family members, had long been considered to be a specifically human characteristic. Yet rats, chimpanzees and other animals also exhibit similar behaviour. Rachel Dale, Friederike Range and colleagues, of the Messerli Research Institute at Vetmeduni Vienna had already shown that dogs also share food rewards with other dogs. Using a bar-pulling task, the dogs delivered the treats to partner dogs – especially if these were already known to them.

A new study by the research team, published in the journal PLoS ONE, now used a more complex task set-up to confirm the prosocial behaviour of dogs. The experiment showed that dogs continued to prefer familiar partners. However, the increased complexity of the task influenced the readiness with which the dogs delivered a food reward to another animal. The study thus confirmed that the chosen method affects the result and is much more dependent on social proximity than had previously been assumed.

Recognition of objects necessary for giving treats

Instead of pulling on a rope, the dogs in the present study had to recognize special objects in the form of tokens in order to deliver a food reward to the other dog. “This time we not only tested a different experimental set-up but also the level of difficulty,” explains Dale. “The dogs were first trained to touch a token in exchange for a food reward for themselves. They were then trained to recognize two more tokens: one that resulted in a reward being delivered to a partner dog and another which did not.” Three experiments were then conducted to test whether the dogs exhibited prosocial behaviour even in this more complex task and whether they would deliver a food reward to a partner or not. The researchers also tested whether it made a difference to the donor dog if the receiver was familiar or a stranger and whether the presence of another dog was enough to trigger generous behaviour in the test dog even if the partner had no access to the food.

prosocial-behaviour-photo-1

For testing prosocial behaviour, donor dogs had to choose via tokens, if they share Food or not. (Photo: Rachel Dale/Vetmeduni Vienna)

Do dogs have to see the recipients to reward them?

The test set-up consisted of two enclosures. The test dog was trained to wait on a specific location in one enclosure until the researchers revealed a board containing the tokens. The dog could then choose to deliver a food reward to the receiver dog or not. In the first test, either a familiar dog or a stranger sat in the receiver enclosure. The dogs could see each other during the experiment. In the second test, the receiver enclosure remained empty but the other dog was present in the testing room. In a third test, the test dogs were alone in the entire set-up. At the end of each test series, the donor animals could reward themselves by being allowed to touch the token that delivered the food reward to them. This was done to ensure that the dogs remained motivated and unstressed and did not become distracted by an unfamiliar dog.

Dogs remain charitable even in complex tasks

The experiment confirmed that dogs continue to exhibit prosocial behaviour despite the more complex task. The dogs clearly showed a preference for sharing the food reward with a familiar dog. Unfamiliar dogs were rewarded nearly three times less often than familiar ones. The higher level of complexity, however, impacted the general frequency of the food delivery. This influence could be shown among dogs for the first time by comparing the token choice experiment with the simpler bar-pulling set-up and confirms the results of similar tests performed with small children and chimpanzees.

prosocial-behaviour-photo-2

The donor dogs shared food even in the more complex token-test, but more likely for dogs they know. (Photo: Mylene Quervel-Chaumette/Vetmeduni Vienna)

Presence of a partner makes dogs more likely to share

The behavioural biologists found another significant difference regarding the question whether the presence of a partner was important for the motivation of the test dog. Even when a second dog was present in the testing room without being in the other enclosure, the donor dogs were more motivated to give a food reward. When the test dogs were alone in the room, the number of food deliveries went down.

This aspect, known as social facilitation, could not be shown in the first study. The social facilitation theory starts from the assumption that animals will perform more readily in the presence of conspecifics. Given a more complex task, the presence of a partner appears to play a greater role. In this case, too, the donor dogs preferred familiar partners. “The difference was smaller, however, than when there was direct visual contact. Social facilitation should therefore be considered and controlled more strongly in future studies and in simple experiments,” says Range.

Source:  Vetmeduni Vienna press release

Gambling wolves take more risks than dogs

Would you rather get 100 euros for certain, or have a fifty-fifty chance of receiving either 200 euros or nothing? Most choose the first, as humans tend to be “risk-averse”, preferring a guaranteed pay-off over the possibility of a greater reward. It is thought that the human preference for “playing it safe” has evolved through natural selection: when you live precariously, like our remote ancestors, losing all your food reserves might be catastrophic, while adding to them might not make much difference to your chances of survival.

Wolf pups

Etu and Ela, two wolf pups at the Wolf Centre Photo credit: Rooobert Bayer

Here, in one of the first studies on risk preferences in animals other than primates, scientists show that wolves are consistently more prone to take risks when gambling for food than dogs. When faced with the choice between an insipid food pellet and a fifty-fifty chance of either tasty meat or an inedible stone, wolves nearly always choose the risky option, whereas dogs are more cautious.

“We compared the propensity to take risks in a foraging context between wolves and dogs that had been raised under the same conditions,” says Sarah Marshall-Pescini, a postdoctoral fellow at the Messerli Research Institute at the Veterinary University of Vienna and the Wolf Science Centre, Ernstbrunn, Austria, the study’s first author. “We found that wolves prefer the risky option significantly more often than dogs. This difference, which seems to be innate, is consistent with the hypothesis that risk preference evolves as a function of ecology.”

The study was done at the Wolf Science Centre, Ernstbrunn, Austria, a research institute where scientists study cognitive and behavioral differences and similarities between wolves and dogs. Here, wolves and dogs live in packs, under near-natural conditions within large enclosures.

Marshall-Pescini let each of 7 wolves and 7 dogs choose 80 times between two upside-down bowls, placed side-by-side on a movable table-top. The animals had been trained to indicate the bowl of their choice with their paw or muzzle, after which they would receive the item that was hidden beneath it.

The researchers had taught the wolves and dogs that beneath the first bowl, the “safe” option, was invariably an insipid food pellet, while beneath the second bowl, the “risky” option, was either an inedible item, a stone, in a random 50% of trials, and high-quality food, such as meat, sausage, or chicken, in the other 50%. As a control, the side for the “safe” and “risky” option changed between trials, but the animals were always shown which side corresponded to which option; whether they would get a stone or high-quality food if they chose the “risky” option was the only unknown. Rigorously designed control trials confirmed that the animals understood this rule, including the element of chance.

Wolves are much more prone to take risks than dogs, show the results. Wolves chose the risky option in 80% of trials, whereas dogs only did so in 58% of trials.

The researchers believe that dogs evolved a more cautious temperament after they underwent an evolutionary shift from their ancestral hunter lifestyle to their present scavenger lifestyle, which happened between 18,000 to 32,000 years ago when humans first domesticated dogs from wolves. Previous research has suggested that species that rely on patchily distributed, uncertain food sources are generally more risk-prone. For example, chimpanzees, which feed on fruit trees and hunt for monkeys, are more risk-prone than bonobos, which rely more on terrestrial vegetation, a temporally and spatially reliable food source.

“Wild wolves hunt large ungulates — a risky strategy, not only because hunts often fail, but also because these prey animals can be dangerous — whereas free-ranging dogs, which make up 80% of the world’s dog population, feed mostly by scavenging on human refuse, a ubiquitous, unlimited resource. So dogs no longer need to take risks when searching for food, and this may have selected for a preference to play if safe,” concludes Marshall-Pescini.

Source:  EurekAlert! media statement

Dogs give friends food (prosocial behavior in dogs)

A readiness to help and a positive attitude toward others are considered foundations of human relationships and human cooperation. But not only humans cooperate and support each other, animals do so, too. A group of ethologists from the Messerli Research Institute at the Vetmeduni Vienna have shown for the first time that dogs also behave prosocially toward others – provided that they know the other dog. The results were published in Nature’s Scientific Reports.

Compared to the rest of the animal kingdom, the human capacity for cooperation is something quite special. Cooperating with one another requires a certain amount of prosocial behaviour. This means helping others without any direct personal benefit.

“Dogs and their nearest relatives, the wolves, exhibit social and cooperative behaviour, so there are grounds to assume that these animals also behave prosocially toward conspecifics. Additionally, over thousands of years of domestication, dogs were selected for special social skills,” explains study director Range. For this reason, Range and her colleagues Mylene Quervel-Chaumette, Rachel Dale and Sarah Marshall-Pescini studied 16 dogs to test their readiness to benefit familiar versus unfamiliar partners.

The researchers studied the prosocial behaviour of the animals using a bar-pulling task in which the dogs had to pull trays and decide whether a second dog would receive a treat or not. In the test, the donor dogs used their mouths to pull a string to bring a tray toward a second dog. They could choose either an empty tray or a tray containing a treat on the partner’s side.

The donor dog (right) can pull a tray and donate food to the receiver-dog (left). (Photo: Mylène Quervel-Chaumette/Vetmeduni Vienna)

The donor dog (right) can pull a tray and donate food to the receiver-dog (left). (Photo: Mylène Quervel-Chaumette/Vetmeduni Vienna)

Dogs donate to familiar partners more often than to unfamiliar ones

Whether the donor dogs knew the recipient made a difference. Donor dogs pulled the giving tray more often for familiar dogs than for unfamiliar ones. “Dogs truly behave prosocially toward other dogs. That had never been experimentally demonstrated before. What we also found was that the degree of familiarity among the dogs further influenced this behaviour. Prosocial behaviour was exhibited less frequently toward unfamiliar dogs than toward familiar ones.

Prosocial behaviour put to the test

In the bar-pulling task, the donor dogs decided whether another dog would receive a treat or not. The donor dog itself did not get the treat. The only purpose of the task was to benefit the other dog. By conducting several control tests, the researchers excluded the possibility that the dogs were simply pulling the trays for the fun of it. Donor dogs were reserved in pulling the tray when an unfamiliar dog was in the next enclosure.

At the end of each test run, the researchers conducted another test to show that the donor dogs knew what pulling the tray meant. They allowed the donor dogs to pull on a tray to give themselves a treat, and all dogs did just that. “This control excludes the possibility that the dogs did not pull on the tray out of fear of the unfamiliar dogs. Given the same situation, the dogs gladly gave themselves a treat,” says Range.

“We were also able to disprove the argument that the dogs pulled the string less frequently because they were distracted by the unfamiliar partner during the test. Only rarely did a donor dog interact with the unfamiliar dog,” Range explains.

Source:   Vetmeduni Vienna media release

Dog-human cooperation is based on social skills of wolves

Dogs are man’s best friend and partner. The origins of this dog-human relationship were subject of a study, published recently in the journal Frontiers in Psychology,  by behavioural scientists from the Messerli Research Institute at the Vetmeduni Vienna and the Wolf Science Center.

They showed that the ancestors of dogs, the wolves, are at least as attentive to members of their species and to humans as dogs are. This social skill did not emerge during domestication, as has been suggested previously, but was already present in wolves. 

Photo by the Wolf Science Center

Photo by the Wolf Science Center

Commonly accepted hypotheses about domestication suggest: “Dogs have become tolerant and attentive as a result of humans actively selecting for these skills during the domestication process in order to make dogs cooperative partners.”

Friederike Range and Zsófia Virányi from the Unit of Comparative Cognition at the Messerli Research Institute question the validity of this view and have developed the “Canine Cooperation Hypothesis”. Their hypothesis states that since wolves already are tolerant, attentive and cooperative, the relationship of wolves to their pack mates could have provided the basis for today’s human-dog relationship. An additional selection, at least for social attentiveness and tolerance, was not necessary during canine domestication.

The researchers believe that wolves are not less socially attentive than dogs. Dogs however cooperate more easily with humans because they more readily accept people as social partners and more easily lose their fear of humans. To test their hypothesis, Range and Virányi examined the social attentiveness and tolerance of wolves and dogs within their packs and toward humans.

Various behavioural tests showed that wolves and dogs have quite similar social skills. Among other things, the researchers tested how well wolves and dogs can find food that has been hidden. Both wolves and dogs used information provided by a human to find the hidden food.

In another study, they showed that wolves followed the gaze of humans. To solve the task, the animals may need to be capable of making a mental representation of the “looker’s” perspective. Wolves can do this quite well. 

Another experiment gave dogs and wolves the chance to observe conspecifics as they opened a box. When it was the observer’s turn to do the same, the wolves proved to be the better imitators, successfully opening the box more often than dogs. “Overall, the tests showed that wolves are very attentive to humans and to each other. Hypotheses which claim that wolves have limited social skills in this respect in comparison to dogs are therefore incorrect,” Range points out. 

At the Wolf Science Center in Ernstbrunn in Lower Austria, Range and Virányi investigated the social behaviour of dogs and wolves that grew up with members of their species and with humans. “The animals are socialized both with conspecifics and with humans. To be able to compare the behaviour of dogs and wolves and to investigate the effects of domestication, it is important that the animals live in the same conditions,” Virányi explains.

Source:  Vetmeduni Vienna media release

Read my other post about the Institute’s research with wolves here

Kathleen Crisley, specialist in dog massage, rehabilitation and nutrition/food therapy, Canine Catering Ltd, Christchurch, New Zealand

Immunotherapy hope

There is a form of cancer treatment called immunotherapy, where antibodies inhibit tumor growth.  Until now, such therapy wasn’t available for dogs.  A research team at Messerli Research Institute of the Vetmeduni Vienna, the Medical University of Vienna and the University of Vienna have now developed antibodies to treat cancer in dogs.

The newly developed antibody brings hope for dogs. (Photo by:  Michael Bernkopf / Vetmeduni Vienna)

The newly developed antibody brings hope for dogs. (Photo by: Michael Bernkopf / Vetmeduni Vienna)

Since cancer cells bear very specific antigens on the surface, the corresponding antibodies bind to these molecules and inhibit tumor growth.  A destructive signal sent by the antibody to the inside of the cancer cell initiates its death. In a second mechanism, the immune system of the patient also destroys the “marked” tumor in a more efficient way.

Josef Singer and Judith Fazekas, both lead authors of the study, discovered that a receptor frequently found on human tumor cells (epidermal growth factor receptor or EGFR) is nearly 100 percent identical with the EGF receptor in dogs. In human medicine EGFR is frequently used as the target of cancer immunotherapy because many cancer cells bear this receptor on their surface. The so-called anti-EGFR antibody binds to cancer cells and thus triggers the destruction of the cells. “Due to the high similarity of the receptor in humans and dogs, this type of therapy should work well in dogs too,” the scientists say. The binding site of the antibody to EGFR in man and dogs differs only in respect of four amino acids.

The head of the study, Professor Erika Jensen-Jarolim, explains as follows: “We expect dogs to tolerate these anti-cancer antibodies well. This will be investigated in clinical studies in the future and is expected to greatly improve the treatment as well as the diagnosis of cancer in dogs.”

The newly developed antibody provides an additional benefit for dogs. As in human medicine, antibodies can be coupled with signal molecules. When the antibody binds to a cancer cell in the organism, the coupled antibody – in this case a radioactive isotope – can be rendered visible and is thus able to show where tumors and even metastases are located. When the selected isotope also contributes to the decay of cancer cells, the approach is known as “theranostics” (therapy and diagnostics).

In veterinary medicine, immunotherapy will be employed for the treatment of mammary ridge cancer (milk line cancer) in dogs. It may also be used as part of a combination therapy.

The team have published their study results in the journal Molecular Cancer Therapeutics.

Source: Vetmeduni Vienna media release