Guest Speaker: Kayce Cover – Training multiple species; Always the Same, Except When it is Different…

I want to introduce you to Kayce Cover who I have met last year at the IMATA conference and long before on social media platforms.

She has trained over 130 different species of animals, ranging from crabs and sharks to dolphins and primates. She has trained animals for diverse applications, including pigs to free stand for a five inch stick into their vena cava for blood draw, cows to tell scientists if they wanted food or a date with a bull, and monkeys to help quadriplegics.

She uses bridge and target technique (reserving free shaping for behaviors such as vocalizing). Augments unconscious learning with conscious learning.  Each learned step is named.  Later, these early learning blocks can be referenced by name, to rebuild a behavior, or build a new behavior.

Uses give rich, immediate feedback and information, using Intermediate and Terminal, bridges, and ‘Name and Explain,’ which is a running narrative of what is happening around, and to, the animal.

Aspires to work fluently, at the ‘speed of communication’.

See her website for more information: www.synalia.com

Training multiple species: Always the same, except when it is different…

When training multiple species, we may find that the mechanics of training are mostly the same, across all species, but the interaction, perception and motivation can differ greatly. Some training skills we often don’t discuss, or even acknowledge, can be very important.  Some of these may be important for us to discuss, and to hone together as trainers develop, in order to produce the best trainers and training, possible. This set of skills can be lumped under “connection”.

The Mechanics that seem to be universal amongst animals:

Operant Conditioning

Operant conditioning provides a fast, sure method of teaching.  From crabs to dolphins and primates, all animals learn very quickly, when the information is presented effectively, and reinforcers are well chosen and deftly delivered. Crab going into a bowl for microscope examination:

The use of targets removes trial and error in training, and can yield clean, speedy results – devoid of collateral behaviors.  Animals are able to quickly, accurately, and reliably produce behaviors on cue, without drift and degradation that can be  seen in free-shaping and luring. Free shaping remains for capturing behaviors which cannot be targeted – like vocalizations, and urination.  Wolf learning basic behaviors and concepts:

Cognitive Teaching

However, it can be useful to take training to another level.  When training animals to perform services for people (like monkeys that aid quadriplegics), it is helpful if the animal understands the purpose of the behavior, and does not rotely perform an action on cue.  For example, if we need a monkey to turn on water at the sink, to fill a glass of water, and the water pressure varies, then the amount the monkey moves the lever may change, day by day, and the monkey needs to understand that we need enough water flow to fill the glass in reasonable time, without water spurting all over the kitchen.  In this video, Name & Explain is used to lead a rhino to cooperate in getting his abscess flushed;  he actually threaded the syringe tube right into his abscess, as had been described for him:

If we teach really simple pieces of behavior, and name each of these, we can mix these pieces, or components, in infinite combinations to create whatever behavior is wanted.  For example, 26 components yielded all the behaviors needed for husbandry, safety, demonstrations, and coping with stress, at the Smithsonian’s National Zoo’s bears, aquatics, and Beaver Valley.  Naming behavior components also makes it easy to break the behavior apart later, if some piece needs repair.  Naming components can dramatically speed training.   Here is an example that led to litter training this horse:

If, as he learns,  the animal observes what he is creating, and learns names for the pieces, there are important benefits gained.  One step further, and it is possible to get feedback directly from the animal to ascertain that he has grasped important concepts.  The animal can be cued using the name, and observed for correct compliance.  He can also be shown to use a toggle to indicate a choice, if we want him to be able to signify back to us that he grasps the particular idea under examination. Then we can confidently move to the next behavior or concept without resorting to great amounts of repetition. Horse sorting tools from foods:

 Things that vary amongst animal types:  Interaction, Perception and Motivation

Interaction:

Animals interact with their environment and one another in diversely different ways.  While orangutans may welcome eye contact, gorillas are different.  Even amongst dogs, eye contact is welcomed by some and irritating to others.  Horses interact with their muzzles, rhinos with their lips and primates with their hands. Turning away can take the pressure off a sloth or a bird, but open you to attack from a tiger.  Some animals will be bolder if you lower your profile, others are likely to attack.  In short, there is no one perfect way to approach, and interact, that is appropriate for all animals.

Particularly amongst dog trainers, there is commonly discussion of reading dogs and how to mimic their social signals to enhance our mutual understanding.  However, whenever we switch between languages, there is always the risk of slipping in the wrong word, and in behavior interchanges, we can easily slip in the wrong gesture.

Sometimes mimicing the animal communication can be helpful (like chuffing to tigers).  But it is good that we can develop an independent system of communication with our animals, one which is the same for all the animals, regardless of species or situation. and based on cues and other labels.  Then trainers have only one communication system to master.

Perception:

Changes in perception between types of animals can confuse us.  We can have training difficulties if we ‘talk’ to the perception an animal does not ‘listen’ to.  It is not helpful to make hand gestures to a blind person.  Scientists tried to train frogs to target something right in front of their eyes, to later learn that the frogs, while staring straight ahead, were not ‘listening’ straight ahead.  They were listening to the changes in the water pressure, which they perceived through their lateral lines!  Pigs do beautifully with scent detections, and not so well with visual discriminations at a distance or if elevated.  When training sea lions to stay in their pools, it was not enough to allow them to stand in the water.  They did not seem to be aware of the water on their skin.  But if asked to float in the water, they easily learned to stay in the pools.

Perception differs between different kinds of animals, but also when an individual moves through environmental changes – as when going from the bright outdoors into a dark barn, or when stepping into water.  Similarly, when teaching a flip,  it can take more time to lead a sea lion to smoothly cross through the air/water interface, than to teach all the other steps of a flip – combined.  When an animal travels through a different medium – air, water, mud – sound, echolocation, vision, conductivity, and other perceptions may significantly change.  It can be helpful to spend enough time at these interfaces, to allow the animals to explore and get used to the differences.

Motivation:

Motivation is the playground of the trainer.  We usually think of starting with primary reinforcers, when training.  These are things the animal innately desires – like food.

However, Initially, our animals may not even like the food we offer them, just because it is unfamiliar. They may not like us, yet.  However, a grey seal may quickly start to eat, if the trainer begins to  leave the instant they play with a fish.  The seals may not initially like the fish, or the trainer, for that matter, but they may like controlling the trainer.  They quickly figure out that they can sometimes control the movement of the trainer by eating their fish.  And the fish begins to be desired.

And so it starts.

Almost immediately, we can create new, secondary associations with things that will later serve to reinforce behaviors we desire.  Games, walks, stroking, puzzles, novelty, play,  praise, verbal excitement, applause – all these can become reinforcers.

Conditioned reinforcers have significant advantages:

1. They are resistant to satiation
2. They build relationship
3. They are private and exist between you and the animal, and help form a special culture between you and the animals.  Anyone can feed an animal, but they will not be able to tap into special games, jokes, and challenges.  I remember teaching dolphins to cross hurdles in opposing directions.  It took great cooperation, precision, and responsibility from both dolphins.  When I first suggested the idea, they dismissed it, and left me alone at the station.  I broke it down into a series of little challenges, which they hesitantly explored – and then aced.  I then conceded their superior abilities… We had FUN.
4. They optimize the exploitation of ANTICIPATION – the most potent access to dopamine!

As we develop rich resources in secondary reinforcers, we can move into a very important aspect of training – teaching animals to change the way they look at things.  We can help our animals dissolve fears, participate in their own vet care, feel confident about transport, and more.  In so doing, we alleviate stress and uncertainty in our animals and help to clear the way for long, healthy, lifespans.  This video shows changing the way pigs look at blood draw, as well as the project where the Intermediate Bridge was developed:

Connection:  Beyond technique and biological differences…..

Is the need to connect with our animals, during our work with them.  It is a constant, calm focus that holds the animal in our presence, as we work.  The connection needs to be supportive, rather than demanding.  We need to ‘see’ the animal’s certain success, rather than urge them to comply or work.  Optimally, the animals will love working with us as much as we love working with them, and they will drive both performance, and behavior development.  Working fluently helps to support this dynamic.

Fluency is delivering cues, feedback, and other information at the natural rate of decoding for a particular animal.

Summary:

Once we understand operant conditioning, we can train any kind of animal.  If we add cognitive training skills, we (and the animals) gain significantly increased benefits.  If we consider, before a training session, the differences in interaction, perception and motivation, we can prepare for a smooth, productive training session especially if we establish, and maintain our emotional contact with our animals.

Enjoy the journey!