Formal Presentations - Thursday
​Abstracts

Watch Videos ​(Login required)​​

EFFECTS OF ENVIRONMENTAL ENRICHMENT DEVICE USAGE IN
REDUCING STEREOTYPICAL BEHAVIOUR OF A SOLO-HOUSED
AMAZONIAN MANATEE (TRICHECHUS INUNGUIS) CALF IN REHABILITATION

Nena Stanekovic

Rainforest Awareness
Education Center
Ontario Veterinary College

Full Abstract

A solo-housed 8-month old Amazonian manatee (Trichechus inunguis) calf housed in a rehabilitation centre since 2 months of age presented with stereotypical behaviour. 3 differing EEDs were provided to the manatee individually from 17:00-18:00 throughout the study periodfor 3 observation periods each at random. The subject was observed for 15 minutes prior to and after EED presentation. 3 observation periods from 16:45-18:15 took place with no enrichment provision as control. The EED provision period showed a decrease in time spent hitting the muzzle against the enclosure and an increase in travelling and exploratory behaviour compared to pre, post, and control observation periods. It is recommended for rehabilitation centres to provide EEDs to manatees to decrease negative effective states and increase activity.

EGG-CELLENT EGG-XITS:
TRAINING MAGELLANIC PENGUINS TO COOPERATE
IN EGG/CHICK CARE BY VOLUNTARILY LEAVING THEIR NESTS

Megan M. Vens-Policky

Shedd Aquarium

Full Abstract

Penguins can be fierce protectors of their nests, so gaining access to eggs and chicks can prove to be difficult. At the same time, monitoring egg development and, hopefully, chick progress is an
important part of husbandry care in a zoological setting. Historically, a trainer would approach a penguin nest, arms and hands protected by wetsuit sleeves and gloves, reach in, lift the penguin, and collect the egg for candling or the chick for weighing. This process was stressful for both birds and trainers. While the trainer's arms were protected, the human and animal bond was not. What would it look like if the birds would voluntarily come off of their nest for reinforcement and allow us to retrieve the egg or chick without nest-related aggression? Could we change the birds' response to our approaching the nest?Through staff buy-in, relationship building, successive approximations, and positive reinforcement our egg retrievals have become much less stressful for both birds and staff. We first conditioned a more positive association with our nest approach and associated tools, clarified which bird we were asking to leave the nest through name recognition, and, finally, built the behavior to both birds in the pair coming off the nest simultaneously.The results speak for themselves. We have adult birds coming to their trainers for both primary and secondary reinforcement while their eggs and chicks are removed from the nest. We are able to safely remove eggs and chicks with less concern of disturbance to the parents and the young. We have seen this protocol work on both established pairs and new pairs. We have been able to build upon previous years' successes. We have been able to transfer the learning to new nesting locations. Our team has increased confidence in their training skills and ability.In conclusion, as this conference continues to show us, our animals can do amazing things if we ask the right questions, give ourselves credit for understanding the science, give our animals credit for their abilities, and all take a brave, first step.

THE GREY AREA OF
PENGUIN SURROGACY
IN THE FORM OF A NEWLY HATCHED PENGUIN CHICK

Hannah Rose

Gulfarium Marine Adventure Park

Full Abstract

Since 2017, Gulfarium Marine Adventure Park has successfully raised 17 African black-footed penguins (Spheniscus demersus) from hatchlings through adulthood. Most of the credit can be given to the imperative monogamous breeding pairs established within our colony. In fact, their consistency in laying resulted in the number of eggs being incubated outnumbering the number of mated pairs within the colony. We realized that if we wanted to maximize the success rate of this endangered species through our breeding program, we had to think outside the box to come up with a solution.

Because African penguins lay up to two eggs per clutch, it only seemed reasonable to expand the available parents that would later rear these eggs to maturation. However, every mated pair in the colony had laid eggs, leaving no other pairs available to raise offspring. The team hypothesized that other individuals within the colony could be encouraged to become surrogate parents for the developing eggs. One of the resident male adult penguins in our collection, South, had displayed naturalistic and instinctual behaviors of egg incubation in the past. This raised the question of could a male penguin raise a chick through surrogacy? South was encouraged to continue displaying the desired rearing behaviors. An example of this was the incubation period of an egg and remaining on the nest during the entire incubation process. These specific behaviors were conditioned utilizing primary reinforcement being placed on the nest. Additionally, the team utilized established relationships with staff members around the nest as secondary reinforcers. Timmy, another male member of the colony, began to show interest in the attention South was receiving on the nest. From there, a mutually beneficial relationship was formed between these two male members of our colony.

The next step was to confirm that these male penguins could become surrogates for a newly hatched chick without endangering the chick in the process. A mock chick was placed on the nest and the males were reinforced for parental care behaviors. This mock chick gave us the ability to test our training without risking the life of a live chick. Their response gave the team the confidence needed to make the decision to present South and Timmy with a live chick that would have otherwise needed to be hand-raised. With the continued support of the training team, South and Timmy successfully raised the chick to adulthood. This unique approach to raising chicks now provides an increase in animal welfare for our colony by way of surrogacy.

CAN I TRAIN A ROBOT
LIKE MY DOLPHIN?
AN OPEN INVITATION FOR THE APPLICATION OF BEHAVIORAL SCIENCE TO
​ARTIFICIAL INTELLIGENCE

Maximilian J. Du

Intelligence through Robotic
Interaction at Scale (IRIS) Lab
at Stanford University

Full Abstract

Operant conditioning techniques developed in zoological facilities have made positive impacts in a large collection of other fields. Traditionally, these impacts have been recognized in the classroom, the home, and the office. Recently, with the rise of more sophisticated AI models,
there is potential for these techniques to play a role in creating robots that learn complicated skills quickly and robustly. Such robots have the potential for broad societal impact, including providing greater independence to older adults and the differently abled. Related developments in AI can also help zoological facilities, including the ongoing efforts in acoustic monitoring of dolphin welfare (Jones et al., 2021). A common type of robot-learning algorithm is focused on maximizing a reward signal. These methods, known as Reinforcement Learning, can be traced back to behavior theory foundations shared with animal trainers, developed by Thorndike and Skinner (Sutton & Barto, 2018). Since then, there have been many examples of convergent evolution between the fields. For example, complex robot behaviors can be shaped by working backward from the goal
(Florensa et al., 2018). This is a behavior chain technique that was originally introduced by Karen Pryor and other trainers (Pryor, 2006). Moreover, roboticists and marine mammal trainers are often restricted to sparse (whistle-bridge) reward signals, which creates a collection of
challenges that have been approached in similar ways by the two fields.
Historically, robots have been orders of magnitude slower to learn than animals. However, recent advances in large AI models have opened up opportunities for robots to learn very quickly. These advances have even enabled the first demonstration of robot “clicker”
training (Mirchandani et al., 2023). As robotics continues down this path, expertise in behavioralscience becomes even more relevant. In this talk, I will give an overview of the intriguing parallels between my field of robotics and the world of marine mammal training. Hopefully, it will be a first step in a rich interdisciplinary discussion that can foster innovations on both sides.