imata logo
Username: Password:  

Forgot Password?
search imata Make a Donation IMATA on Facebook
Masthead Image

Home / Membership /

Sink and Swim

By Barbara A. Losch, Disney's Animals, Science and Environment, The Seas, Epcot, Walt Disney World Resort

Originally published in Soundings Volume 41, Number 2 — Second Quarter 2016

Located in Epcot at Walt Disney World® Resort, The Seas is a 21.6 million liters (5.7 million gallon) re-creation of a reef community. It spans 61.9 meters (203 feet) in diameter and is 8.2 meters (27 feet) deep. It is home to 57 different species with 1,190 individuals, including teleosts, elasmobranchs, sea turtles, and dolphins. The Seas has no above-water viewing, but underwater windows surround the perimeter of the facility on three sides allowing guests continuous opportunities to look into the habitat. The dolphin area encompasses 25 percent of the total exhibit and has 18 windows—one row of 12 at 2.4 meters (8 feet) below the surface and one row of six at 7.3 meters (24 feet) below the surface—for viewing our four male Atlantic bottlenose dolphins (Tursiops truncatus).

Photo 1:  Dolphin Habitat

Scheduled dolphin presentations occur in a center observation area where a narrator describes our research or provides general dolphin information in front of two large underwater windows. (For details about our cognitive research with dolphins, please see Harley, Fellner, & Stamper, 2010.) Typically during these presentations, one dolphin is interacting in a research session in front of the guests while the other three dolphins are involved with training at other locations within the habitat. Since the upper windows are situated 2.4 meters (8 feet) below the surface, most guest observations of the non-session dolphins were of the lower half of the dolphins’ bodies. To increase dolphin visibility during sessions, the trainers decided to teach a sink behavior in order to give guests a better view of the dolphins while they performed their behaviors. Training the behavior included the following steps: 1) training the dolphins to descend or sink, 2) pairing the sink with other behaviors, 3) maintaining depth during behaviors, including perimeter behaviors, and 4) keeping the dolphins focused on the trainers rather than the guests.

Training the sink

Training started with Calvin and Malabar, 21 and 15-years-old, respectively. Both knew a spy hop or bob behavior, where they pop up out of the water as their trainer jumps up and sink slightly underwater when their trainer crouches down. We started delaying our jump-up after the dolphins sank in hopes that our delay might cause them to pause before popping up to the surface. Both dolphins delayed with this tactic and stayed underwater briefly. We bridged that delay. We were able to lengthen the time they spent underwater by bridging for longer and longer delays. Pairing a new sink SD with the spy hop SD during this phase of the training enabled us to fade out the spy hop SD in just a few sessions.

Our two other dolphins, Khyber, 23, and Ranier, mid 30s, had not been taught the spy hop behavior. Neither dolphin knew an existing behavior involving a sink. We decided to use a target pole to teach them to descend. Placing a long target in the water, we pointed to the target and bridged as the dolphin descended, making sure the bridge occurred on the descent, not as they reached the target. We again added the sink SD during the target training and slowly faded the target out of the picture. Eventually, all the dolphins were sinking on the new  SD.

Pairing the sink with other behaviors sinking

Since we wanted the dolphins to do multiple behaviors on a single sink, we started incorporating an up signal. We started by pairing an up SD with the bridge after a descent so that they would learn to wait until they saw an up SD before surfacing. Once they learned to wait for the up SD we started asking for behaviors while they were submerged.  Adding behaviors while they were under water was fairly easy. If the dolphins accidentally surfaced on a bridge for a behavior we would just ask them to go back under, ask for a behavior, bridge and then quickly give the up SD. Soon they were doing several behaviors and remaining underwater even after hearing multiple bridges. The meaning of the bridge transitioned from communicating that the sink behavior was correct and complete to communicating that the sink behavior was correct but you may need to stay under water for other signals.  During this portion of the training, we did encounter a problem. The dolphins were drifting upward while they were doing their behaviors, so much so they ended up just under the surface of the water after a few behaviors.

Photo 2: Guest Viewing

Maintaining depth

We tried a variety of techniques to prevent the upward drift from occurring. The first was to ask for the sink in a shallow area of 1.8 meters (6 feet) within their environment. We then asked for behaviors while they were touching the bottom and made sure they stayed at 1.8 meters (6 feet) during a multitude of behaviors. We hoped the concept of depth would carry over when we asked for sink in our deeper environment. Unfortunately, the depth concept did not carry over. It seemed we had made the body position more important. When we asked them to sink in the deeper zone, they went down 0.6 meters (2 feet) and positioned as if they were touching the bottom of the 1.8 meter (6 foot) area. Looking back on the training, this made perfect sense. Most of our previously trained behaviors are all about the dolphins’ body position and not where they were in the water column.

During sink training, we noticed the dolphins would fly up out of the water after a sink, at times so high their flukes were almost out of the water.  Capitalizing on the height from their excitement, we quickly gave the SD to sink and the momentum would carry them generally a foot deeper in the water column.  Using selective bridging, we were able to capture deeper descents. We also discovered that once the dolphins were submerged, if we kept our hands out over the surface of the water, they would maintain their depth.   We then gradually added behaviors  and worked on maintaining depth.. Over time, all the dolphins learned to reposition themselves to a lower level if they happened to drift upward during a behavior.

Photo 3: Calvin and Malabar just under the surface of the water

Perimeter behaviors

Pairing sink with some of our existing perimeter behaviors was another goal in the training process. Our perimeter behaviors involved the dolphins swimming at the surface around the edges of their environment while maintaining a specific behavior. With the sink perimeter behaviors, the dolphins would have to do their behaviors while maintaining a depth of 2.4 meters (8 feet). We used two trainers, one at the surface and one at the window. A dolphin was asked to go to the window trainer; the window trainer then asked the dolphin for one of the perimeter behaviors and moved with the dolphin past the twelve windows, attempting to keep the dolphin at an 2.4 meter (8 foot) depth. There are two corners on the path the trainer would take past the windows. One of the corners leads from three windows into a long corridor of six windows. The other corner leads from the corridor to the two presentation windows. Both areas had an area where the trainer would be out of view of the dolphins.  During the first few attempts around the first corner, we lost the dolphins.  Back chaining the route was the solution to this problem. We would start training the route from the end and work our way back to the start.

The end would be at our two presentation windows. All of the dolphins were familiar with these two windows as the majority of our presentations occur in this area.  The dolphins already knew to maintain behaviors at window level while going from one window to the next.  Now we needed to move around the corner to the corridor.  We started at the window just past the corner. 

We gave an SD and then ran around the corner to the first presentation window while continuing the SD. The dolphins initially stopped after they rounded the corner but saw the trainer continuing the SD and continued past these windows with little need of assistance. Once they were comfortable making the corner between the corridor and the presentation windows, we easily added two or three windows at a time to the route and worked our way along the corridor. We faded ourselves out by stepping further and further away from the windows and soon they were traveling the route without us moving with them.

When we added the second corner at the top of the corridor, the dolphins got around it easily. We added more and more windows and soon the dolphins were able to do the entire route without the trainer with them. We then faded out the window trainer. The trainer topside could ask for a sink and send the dolphins on perimeter behaviors without the assistance of a trainer down at the windows. On a side note, throughout the perimeter training, we alternated behaviors as we trained the route. This generalized the route and enabled us to pair the sink behavior with any of our previously trained perimeter behaviors.

Photo 4: Window trainer moving past windows on route

Perimeter tumble

One behavior we included in our perimeter behaviors was an underwater tumble or somersault. The dolphins could do a stationary tumble in the windows but had not learned to travel while they were tumbling. They tumbled on a sink SD but remained in the same location. We worked the perimeter tumble three different ways.  Which technique we used depended on staffing strengths and progression of the behavior.

  1. We used a second trainer at the window just as we had when training the other perimeter behaviors. We gave additional tumble SDs at each window.
  2. The dolphins were already tumbling on a sink, and they would do A to Bs on a sink. We first positioned trainers along the top of our environment. Then a trainer on one of the training docks would ask the dolphins to sink and send them to a person positioned above the first window. This trainer would give the tumble signal and send them to the next person and so on.
  3. We moved one of our floating docks near three of the windows, positioned it out a bit so a dolphin could fit between the window and the dock and then asked the dolphin to sink and tumble while the trainer walked above them giving additional SDs.

Using a combination of the above techniques, two of our dolphins are currently doing perimeter tumbles along the entire route. 

Photo 5: Guests watching Calvin

Trainer focus vs. guest focus

Another challenge we faced was to keep the dolphins’ focus on their trainers while guests were attempting to interact with them on the other side of the window. During the initial training stages, we worked on sink at numerous locations: our off-exhibit holding areas, all of our dock locations, standing in the water,, and from our catwalk that surrounds the dolphin area. We worked on sink when there were lots of things going on around the dolphins and gave large reinforcements for attention during distractions. This technique seemed to work as we moved to the windows. The dolphins stayed trainer-focused even with guests waving at them from the other side of the windows.

Photo 6: Guests viewing dolphins’ lower bodies only

The sink behavior has been great fun to train, and we continue to incorporate different aspects of this versatile behavior. Guests’ views of the dolphins have improved. We went from having only the lower half of a dolphin’s body on view to giving visitors a full body perspective of the animals during the sessions. Our next goal will be to further the descent behaviors to our lowest windows, which are 7.3 meters (24 feet) below the surface of the water.  Currently, two of our dolphins are descending to 7.3 meters and we are just starting to ask for behaviors while they are at that depth.

Photo 7: Guests with full view of dolphins



Harley, H.E., Fellner, W., & Stamper, M.A. (2010). Cognitive research with dolphins (Tursiops truncatus) at Disney’s The Seas: A program for enrichment, science, education, and conservation. International Journal of Comparative Psychology, 23,331-343.