Sunday, March 23, 2014

Training 101

Courtesy of SeaWorld
What is operant conditioning? It is the reinforcement of conscious behavior deliberately offered by the learner. Huh? A person or animal demonstrates a behavior he or she has learned to get a reward or avoid punishment. Telling your dog "good boy" and patting him on the head when he sits is an example of operant conditioning. So is the training SeaWorld keepers use to get marine mammals to perform complicated maneuvers. So is Sheldon training Penny on "The Big Bang Theory."

 

Zoos, aquariums, and companion animal trainers use operant conditioning techniques to teach animals behaviors and shape those behaviors for husbandry, veterinary, service, or entertainment purposes. When teaching a new behavior, trainers try to remember each animal's biological constraints. For example, solitary animals might not perceive praise and attention as a reward. So a reward that will work for a dog might not work for a tiger. And many animals have physical constraints. It would be nearly impossible to train a tortoise to jump through a flaming hoop!

Training Xiao-Li, the male red panda
at the Binghamton Zoo. Photo courtesy of Saasha Caldera.
Through repetition and successive incremental steps, animals learn to perform various behaviors on cue. A cue could be a verbal command, hand signal, sound, or object....basically any stimulus that the animal has associated with a particular behavior. Once the animal performs the desired step or behavior correctly, the trainer reinforces the behavior. There are two types of reinforcement: positive and negative. Positive reinforcement is something the animal wants: food, praise, petting, etc. Negative reinforcement is something the animal wants to avoid: a squirt of water, a swat, an annoying sound. Positive reinforcement has proven to be the most effective type of training for both people and animals, and is normally used in zoos and aquariums. Have you ever heard the saying, "you can train a donkey using a carrot or a stick"? Zoos and aquariums use the carrot.

What if the trainer can't give the positive reinforcement to the animal right away? Trainers want the animal to know which exact behavior is being reinforced, and they don't want the animal to become frustrated if there is a delay between the behavior and the reinforcement. Timing is everything! So trainers use a "bridge" to fill the period of time between the behavior and the reward. A bridge can be verbal (saying "good"), or it can be a clicker, or a whistle. The bridge means, "You're done, good job! A reward will be coming any second." After the animal becomes habituated to the bridge, the bridge actually becomes a secondary reinforcement. In this way, trainers can string together multiple behaviors, rewarding each behavior with a bridge, and give some type of positive reinforcement at the end of the series.

The photographer was more interesting
than I was! Courtesy of Saasha Caldera.
What if the animal ignores the trainer or doesn't want to train? A trainer never chases after an animal, forces it to train, or bribes it to behave. If the animal gets distracted in the middle of the session, the trainer will ignore the animal until the animal decides it wants to train again. The trainer's attention is positive reinforcement for wanting to train. If an animal does a behavior exceptionally well, that animal will get a "jackpot." A jackpot is an extra-big reward to make the animal think, "Wow! That was awesome! How can I get that much fish again??" So a jackpot is a very memorable positive reinforcement that increases the likelihood that the animal will perform the behavior again. At the end of each session, every trainer tries to end on a good note to reinforce his or her bond with the animal.

I've been lucky enough to train a variety of animals in my career so far. I've trained raptors at the Vermont Institute of Natural Science, and at the Binghamton Zoo, I trained Xiao-Li, the male red panda, and Leroy and Tala, male and female North American river otters. The VINS raptors were trained to fly from handler to handler or from perch to perch. The Harris's hawk was even trained to catch a fake rabbit lure on command! I didn't train any new behaviors with these birds; instead, I reinforced behaviors that were previously trained. At the Binghamton Zoo, I trained Xiao to touch his nose to a target stick, which is an easy way to move an animal voluntarily from point A to point B. I also trained him to stand on a log station so a vet could give him a physical while he was awake. I even trained him to voluntarily accept vaccinations and go into a crate on command!

In this video, I am training Tala, the female river otter at the Binghamton Zoo. Both otters are now trained to touch their noses to a target stick and hold in place, squeeze themselves into a PVC pipe attached to the door for voluntary injections, open their mouths on command, shift into different holding boxes, and stand on a scale. I trained the otters for two years, and they were very fast learners!

Elise Newman trains "Tala," the Binghamton Zoo's female North American river otter, to perform behaviors on command, which simplifies husbandry and veterinary procedures. You can hear the clicker "bridge" after each behavior before she gets rewarded with tasty fish pieces! In recent news, Tala (aka "Elaine") just had triplets at the Binghamton Zoo!


One of my favorite parts of training is the "lightbulb moment" each animal gets when it figures out a new behavior. I also love the bond that I've formed with each of the animals I've trained. I can't wait to increase my experience and train more animals!

No comments:

Post a Comment