A Big Step for Carson

I have taken Carson’s training nice and slow. I go at her pace and let her tell me what she is comfortable with. But sometimes I have to take a risk. When Carson was inside last week to escape the cold temperatures outside, I decided to leap ahead in my carefully sequenced training plan to see how she would do in front of a group of people. We have spent months building trust in the mew, but I didn’t know if that comfort level would transfer to the classroom.

I was relieved to have a small group of well-behaved adults for Saturday’s “Talon Talk” program; it would be the perfect audience for Carson’s latest debut. Her first test was coming out of the crate. Just like we had practiced over the previous week, she hesitated at first, looking between the glove and the room outside the crate door. After a moment she set one foot on glove and tested its stability. My hand held her weight without wavering so she peeled the second foot off the perch and placed it on the glove.

Her next challenge was staying calm in front of the audience. I watched her body language out of the corner of my eye. She stood up straight, looking around at each corner of the room, and even took a few mouse tidbits. I was impressed; she didn’t seem concerned about a thing. After a few minutes, I brought her back to the crate while she was still comfortable, completing her portion of the program as a fun, positive experience for both of us.

Inside a Training Session: Not Interested

Sometimes training sessions go really well with big breakthroughs like the session I described with Theo last week. But sometimes we just don’t get anywhere. Here is an example of another session with Theo that was not as productive:

I stood outside and peered into the mew. His yellow eyes almost seemed to glow as he stared back at me cautiously. After making sure I was staying put, he turned away to watch the backyard. I silently willed him to turn his attention to the chunk of mouse sitting next to him on the perch. My mind controls proved ineffective, however, as something else caught his eye. He bobbed his head down, then stretched every one of his fourteen neck vertebrae to capacity as he reached his head to the left. Eyes wide, unblinking. I also swiveled my head around, searching for what would be so interesting to an owl. Did he just now notice the two men sitting quietly by the library? Otherwise the backyard was quite peaceful. Perhaps it was something inside the mew. Would a rogue fly warrant such focused attention? Finding nothing alarming myself, and Theo’s neck still stretched out as far as it would go, I tried to gently refocus his attention with more than mental vibes. I shuffled my feet, quietly crunching the gravel to remind him that I was still there and this was a training session. No luck. His eyes were still round globes locked on something more interesting than me.

Inside a training session: Theo learns the rules of the game

Theo’s training log helps me keep track of his progress.

I wait patiently just outside Theo’s open door. To avoid a direct gaze that could be intimidating, I stare at the ground and count in my head: 34…35…36. In my peripheral vision I can see Theo completely ignoring the two pieces of quail meat sitting on the perch two feet away. He simply rests with one foot tucked deep into the fluffy feathers on his belly. 86…87…88. I am starting to get worried. All he has to do is walk over and eat the tidbits, but his window of opportunity, 1 minute and 45 seconds, is closing in fast. Unfortunately he seems quite content to keep watch over the quiet backyard. As my internal count reaches 105 seconds, I sigh and take a step forward. Theo instantly turns to me. I reach out with the forceps to take one tidbit from the perch. Theo’s eyes lock on the retreating meat and follow it all the way back to the pouch hanging from my belt. Before I can move again Theo places both feet firmly on the perch and his head begins to swivel quickly back and forth between me and the single tidbit left on the perch. He knows I am about to take it away. I pause. Finally he makes his move and walks down the perch to grab his food. I smile at this bold move and give him more quail as a reward. I marvel at how much this owl has learned in the past few weeks. He has learned enough about me to actually predict my behaviors and act to change them. Owls may not be known for their intelligence, but they certainly are not “bird brains!”

The Mouse Went Down the Hole

Imagine a hawk sitting on a fence post along the edge of a cornfield. Scanning the ground beneath her, she suddenly spots a mouse and must decide whether she will chase it or not. There are several factors that influence her decision. First, is she even hungry? If she just finished eating a rabbit, she probably won’t feel the need to work for an extra snack. She also considers if it is a desirable piece of food, whether the tiny mouse is worth her effort, and if she can obtain it without injury to herself. Now if she takes too long to consider all of this, she runs the risk of the opportunity disappearing completely: the mouse could scurry down a hole, out of reach. She will learn to think faster next time!

This “mouse went down the hole” theory is a useful tool in training our raptors. Knowing that their food may disappear, they are more motivated to act quickly. While station training Theo, I placed a piece of meat on a particular perch and gave him 5 minutes (his “window of opportunity”) to come eat it. If he didn’t move by the end of the 5 minutes, I took the food away, just like the mouse escaping down the hole. He learns that he must move before the 5 minutes are up if he wants the tidbit. I gradually shorten his window, working it down from 5 minutes, to 4 minutes, and eventually to a few seconds. I am happy to report that he is currently coming to his station as soon as I set the food down. In just a few weeks, he decided that he is not going to let those mice slip away!

Photo by Vera Domingues/Hopi Hoekstra. 

The Bridge

The bridge, usually a sound like a click or whistle, is an important part of training with positive reinforcement. You may already be familiar with bridges, especially if you have attended a dog obedience class. Metal clicker boxes have become a popular way of training dogs: ask the dog to “sit,” and as soon as his butt hits the ground, you click and give a treat. The click marks the sit behavior, essentially telling the dog, “THAT is exactly what I want you to do, a treat is coming!” It is more precise than using treats alone. After sitting, the dog might stand up by the time he eats the treat; without a click, the dog might think the reward was for standing. The clicker acts as a bridge, appropriately named because it bridges the gap of time between the behavior you want (sitting) and when they get the reward.

The bridge can be any sound or symbol the animal learns to associate with food. I simply make a “cluck” sound with my tongue for our birds. In this short video, I use the bridge to teach Carson to touch the target stick. Now that you know what the bridge means, watch the video carefully to answer these questions:

• What mistake do I make with the bridge when Carson touches the target stick?
• How can you tell that she understands what the bridge means?

Could you spot my mistake? In this session, multitasking with my camera divided my attention and I actually bridged AFTER she touched the target, a second too late. If she didn’t already understand the target behavior, a late bridge could have confused her even more. Did you notice how she responded to the bridge? Looking around in different directions, she searched for the food she knew she earned. Rest assured that after I set the camera down, she got a juicy reward!

Reading body language

Examples of stress, comfort, and ultimate relaxation in great horned owls.

Body language is everything. It tells me when Theo is comfortable with the situation. It also tells me when I have pushed him too far past his comfort zone, potentially deteriorating the trust between us. When it comes to reading these stress signals, owls tend to be very subtle. A kestrel might alarm-call loudly and fly away in a stressful situation, but an owl might stand perfectly still and simply wink one eye. They may rely on blending into the background to avoid confrontations; great horned owls will raise the tufts of feathers on top of their head to enhance their camouflage. They will even hiss and clack their beak if the threat continues. These three responses can be seen in the example photo on the left: winking eye, feather tufts raised, and beak open in a hiss. This is not a happy owl. When I see these signs in a training session with Theo, I listen. I slow down to make him feel more comfortable.

Now contrast that with the owl in the center photo. He is standing comfortably with feather tufts lowered and is watching something in the distance, not the photographer. He is calm and not worried about anything around him. When Theo is not interested in training, he will often turn away and look outside. Ignoring me is one of the biggest signs of trust he can give. One behavior I am still waiting to see with Theo is a stretch, like the one pictured on the right. Stretches and preening are comfort behaviors that can make the bird more vulnerable, so they are only done when it feels extremely safe. When I see Theo reach out a wing and a foot, I’ll know that I’m okay in his book.

Inside a training session: Carson’s “Ah ha!” moment

I hold the target stick about one foot away from Carson. She steps over to stand next to it. In our previous training sessions, I would have rewarded her for that movement. But she has been doing so well moving toward the target that she is ready for the next step: actually touching it. With no way to simply explain what I need her to do, I can only wait. She looks between me and the target a few times before stretching her head forward. I quickly give her a treat. Even though she didn’t touch it, I want to tell her she is going in the right direction. Now her interest is piqued and I can practically see the gears turning in her head: how do I earn another chunk of rat? Her eyes focus on the target as she tips her head to one side. She realizes this strange stick is the key to earning her food, but she doesn’t know what to do with it. She takes a half-step closer and looks up at me hopefully: is this worth a treat? I don’t respond so she looks back to the target and bobs her head. After a moment she moves forward and nips at the target. Contact! I offer her a heap of tidbits as her jackpot prize. I am overjoyed to see this breakthrough and Carson is just as excited to understand with her “Aha!” moment.

Positive Reinforcement

Aldo inspects a bite-size quail tidbit offered on the tip of my finger.

There are many ways to teach an animal to do a single behavior, like having a raptor stand on a gloved hand. Falconers use a traditional method called “manning” where the falconer sits with a new bird on glove for an extended period of time with the idea that it will become acclimated after hours of exposure. A more modern approach to training the same behavior uses positive reinforcement.

Positive reinforcement is everywhere in our lives. You lavish your dog with attention if he comes when called. You receive compliments for a new hairstyle. Both of these examples have a reward that makes the behavior more likely to occur again. Your dog will keep running to you if he loves the attention, or you might style your hair the same way to get more compliments. I use the same concept with our raptors, using food as the main reinforcer. Aldo gets a meat tidbit every time he steps on glove or stands still on the scale. New stimuli, like large groups of people, are introduced slowly. Rather than expecting the bird to quickly acclimate, we use food to turn it into a positive environment for the bird. Both manning and positive reinforcement are proven effective, but the latter is favored by education raptor trainers.

New training ideas

Dan (naturalist at The Raptor Center) practices targeting in a training session with “Strix” the Barred Owl.

At The Raptor Center two weeks ago, Aldo and I learned about more than beak trimming. I had the opportunity to talk with naturalists Gail and Dan about raptor training. It is always good to brainstorm with people that might have new ideas and different backgrounds. Gail and Dan proposed lots of ideas for our birds, like target training. “Targeting” is a common behavior used in zoos where the animal learns to touch its nose to an object, usually a tennis ball at the end of a dowel. It is useful to guide the animal around the enclosure or to teach new behaviors. I have targeted with monkeys, rabbits, and tigers, but hadn’t thought to try it with raptors. When we returned to the museum, I used a stretch of Vetrap to transform an ordinary wooden dowel into a target stick modeled after the one The Raptor Center uses. I introduced the new target to Carson (Red-tailed Hawk) and within one week she was eagerly taking food held next to the target. Soon she will learn to touch her beak to the end of the dowel to earn her treat. New behaviors like this keep training interesting and fun for trainer and animal alike.

Two-way Communication

Carson quickly learned where her training station is and will often wait there for the next training session!

When I first met Carson (our red-tailed hawk), I raised my leather handling glove to her feet and she hopped away to the other side of the mew.  That’s one of the unique challenges of a living collection: the bird can choose whether she wants to participate or not.  Carson was telling me that she was not comfortable stepping on my glove and immediately the gears in my training brain started turning.  Why did she avoid my glove?  How can I make the glove a positive place?  Operant conditioning, positive reinforcement, bridging, and shaping discussed at length in my books and papers on animal training hold most of the answers to these questions.  I decided to start from a different place, however.  I’ve found that for all of these technical theories to work, you need one simple thing: two-way communication.  I was already listening to Carson’s body language, which told me I needed to back up in our training.  I also need to be consistent with my actions so she knows what to expect.  If she steps on the glove, I give her a piece of food; if she does not step on glove, I leave the mew along with her opportunity to earn treats.  When I am consistent and respect her body language, she gains the power to control the outcomes and will work harder to earn the positive ones.  We established a training station, a specific perch where we train.  This allows her to choose if she wants to participate in the training session.  Training with respect to the bird’s comfort level may take longer than other methods, but it is less stressful for the bird and we will build a strong relationship.