Cows in HundCampus

Dagens innlegg er en gjesteblogg fra vår forskervenn Carola, som har vært på Hundcampus flere ganger.

—————————————————————————————————–

My name is Carola Fischer-Tenhagen and I work at the Freie Universität
Berlin in the veterinary clinic for animal reproduction. A year ago I had the idea, when dogs are able to smell cancer, diabetes, drugs or explosives, they must be able to smell, when a cow is in heat as well.

This is a big challenge in commercial dairy farming, because fertility of cows is closely linked with the right moment of insemination. And cows only produce a lot of milk, if they have a new calf every year.

As a first step I collected vaginal fluid, milk and urine of cows in heat and not in heat. I heard from a Norwegian colleque of HundCampus, that here you can train any kind of special search. And so Maxi, my 5 year old Labrador bitch, and I travelled in March 2010 to HundCampus for the first time. It seemed to be very easy for the dogs to identify and differentiate samples from cows in oestrus and dioestrus. First we tested on vaginal fluid and then on milk and urine. After only 50 contact with the target sample (oestrus) the dogs could identify new samples with an accuracy of 80 %.Home again, I realized that it would be much better, if the dog could smell oestrus from the front of the cow, so the danger of being kicked or shit on is not so high. I read about an Indian science group, which proved that rats were able to smell oestrus in the saliva of cows. I thought, if rats can, dogs can as well.So I collected saliva samples of cows in heat and not in heat and came back to hundcamppus.  This time I took Steffi along, a doctoral student in our institute. My hope was, that dogs could equally fast be trained to identify oestrus in saliva, than in vaginal fluid. So we standardized training time to 50 positive reinforced contacts to saliva of various cows in oestrus. We had five dogs in the training; all of them were trained on vaginal fluid as well. After this we did a blinded test, if the dogs could really identify oestrus. Results were not satisfying. Most dogs had only one correct out of four possible, this is the same as pointing by chance.But we learned a lot out of this. All dog handlers were very nervous on the test day, and they behaved totally different for their dogs. In addition to this, dogs were rewarded for indicating the wrong sample, this confused them very much. We stopped this experiment, and retrained the dogs on the testplattform, where dogs are always confronted with a new sample, and they do not have the chance of learning the individual smell of a cow.We also trained the situation that the dog handler do not know the position of the target sample, so he can not give hidden clues to his dog. Before rewarding he was told the correct position, so the dog was not rewarded wrongly.Tomorrow is the next test day, and we are all very excited for the new results. Only the dogs, they seem to be very tired of all this intensive training.

Carola Fisher-Tenhagen