#fish training

Turn off her filter, and she would still get stress stripes when her tank light was on. Turn her tank light off, she would be spooked every time one of the plants in her aquarium touched her. A stranger sitting in front of her aquarium resulted in clamped fins and a ceasing of all movement on her part. Move anything even slightly out of place during a water change without moving it back again- you guessed it, clamped fins and stress stripes. Being constantly stressed out is obviously not a good way for an animal to live, so I started working on counterconditioning and desensitization training with her almost immediately.

We started with the things that were most important to her long-term health and well-being; primarily, her filter. I swapped out the HOB filter my mom had set up with a sponge filter, setting it up with a gang valve so I could exercise precise control over the strength of the water current coming out of the filter. We started with barely any flow at all, though that was still enough that Sophie got stress stripes anytime the filter was on. She was fed all her meals in the back of her tank, progressively moving closer to the filter each time, until eventually all her food was dropped directly above it. It took a couple of weeks, but eventually, she was eagerly swimming over to that spot without clamping her fins, losing any color, or displaying stress stripes. Slowly we upped the strength of the filter, using the same process every time, and now it can run at full strength and she won’t react in the slightest.

We followed the same general process with having her lights on, introducing new things to her tank, leaving new objects sitting outside of her tank, and so on. Interestingly, the more we worked on these things, the easier it became, and the faster Sophie started reacting positively to the stimulus instead of negatively as she had before. I vividly remember setting a coffee cup in front of her aquarium one morning several months after she first came home, and watching her swim up and wiggle at it, despite never having done any training with cups before. I was stunned, but also incredibly proud of her!

Fast forward to two days ago, I removed one of Sophie’s old plants from her tank to replace it with a new one. At first, she swam over to the plant, looking at it with that familiar clamped-fin look and slightly dimmer colors (she virtually never gets stress stripes anymore, thankfully, but will get a little pale if something is stressing her out). All it took was one meal scattered throughout the leaves of that plant, and she no longer reacts negatively to its presence at all. I’ve even caught her sitting on it multiple times these past few days! This is something that would have taken weeks to accomplish when she first came home, and I can’t help but be proud of her for how much progress she’s made these past 8 months.

Probably the most interesting thing, however, has been applying these training techniques to the presence of an unfamiliar person. It’s not me anthropomorphizing- she genuinely reacts differently to someone she doesn’t recognize, as evidenced by the way she stops moving, clamps her fins, and just stares at them as opposed to her normal darting about and wiggling. My mom described this when she first told me about Sophie's behavior around her, and I have seen it myself in the months since. I had people feed Sophie whenever they came over/into my room, and though she will still stop and stare at new people, it only takes a couple of visits now for her grow comfortable enough to actually swim around when they are watching her, as opposed to the weeks it took before.

Interestingly, Sophie never actually reacted this way to me. The very first time I sat down in front of her she swam right over to me, made eye contact and wiggled, and has been just as open and interactive with me ever since. I couldn’t tell you why, but I would guess that maybe, because I have spent more time with fishes (and particularly fishes like Sophie) than most people who have met her, I approached her in a way that was more encouraging of her coming up to me. Who really knows, though! It will always be a mystery what exactly she was experiencing on her end when we first met.

This turned out kind of long and rambly, but it’s just something that’s been on my mind a lot these past couple of days. At times I feel some people do not really understand that fish behavior can be just as complex as that of other animals, and that, sometimes, they have issues like this that won't just magically go away on their own. Sometimes, fish keepers have to do more than just ensure water parameters are good and the diet is appropriate. I think Sophie’s case is a good example of this, and I certainly have learned a lot from working with and getting to know her.

The training I do with the bettas is entirely (aside fromone or two rare occasions) based in positive reinforcement. If you aren’t alreadyaware, positive reinforcement can basically be defined as this: adding astimulus (‘positive’) with the intent of increasing the occurrence of a desiredbehavior (‘reinforcement’). In this context you are, essentially, giving theanimal something they want (i.e. food) in exchange for them performing abehavior you want them to (i.e. pushing a ball). That thing they want canbe anything they find reinforcing (essentially, anything they like and willwork for). For my bettas, this is usually either food or social interaction/attentionfrom myself (yes, this does work- some bettas find attention very reinforcing).

To train the ball-pushing behavior, I used a trainingtechnique called shaping, or more specifically, micro-shaping. As defined bytrainer Alexandra Kurland, micro-shaping is “reinforcing the smallest try” theanimal makes towards the desired behavior. So basically, in teaching Duo tofetch a ball, I am reinforcing (or rewarding) him for every small step he makestowards fetching the ball. When he is reliably performing one step, I then moveon to the next, and so on. Using this technique allows you to break thetraining process down into smaller, more manageable steps, and I have found itto be very effective for a behavior like this that, in my experience, is more challengingfor a domestic betta to learn.

To start, I took the end behavior- fetching a ball- and broke it down into all the small stepsthe fish has to move through in order to perform the whole behavior. That mightlook something like this:

Look at the ball

Swim towards the ball

Touch the ball

Push the ball in my direction multiple times

Leave the ball

Not every fish will follow this progression exactly, andsome will skip steps while others will require additional ones. With Duo, Istarted by reinforcing him for turning in thedirection of the ball whenever I placed it in his aquarium (by giving him food at the exact moment he did it). Soon, he was quicklygiving the ball his attention every time I placed it in his aquarium, so Imoved on to only reinforcing him once he started swimming in the direction ofthe ball. At this point, I had to start paying really close attention to hisbody language so that I could really reinforce the “smallest try” he wasmaking towards performing the next step (biting/pushing the ball).

Duo did not immediately go from swimming towards the ball tobiting it; instead, as he came near the ball, I could see how his body languagewould change in a way that suggested he might be working himself up to bitingit. It’s a little hard to put into words, but if you’ve ever seen the way a bettacarries themselves when they’re thinking of biting something, that’s what I’mtalking about. The way they look at something, the angle of their body in relation to the object, howtheir fins move; those are all things you can (and should) pay attention to here. For a comparison,if you were teaching a dog to bark on cue, you might reinforce them at thispoint for the changes in their facial muscles that occur just before they actuallyopen their mouth to bark, or for opening their mouth even if no sound came out. So I started reinforcing Duo for that change in bodylanguage, and before long, he was using his mouth to push the ball around.

Next, I started working on getting him to push the ball multipletimes before being reinforced, and on getting him to push the ball with greaterforce. This was primarily achieved through using high-value reinforcers (thingshe finds more reinforcing and will work harder for). For Duo, these have beenhis favorite foods (such as Bug Bites pellets, crickets, crab meat, and canned mealworms).Using these reinforcers, Duo naturally started to push the ball more times in arow and with more force. I didn’t really have to spend time breaking thingsdown here; instead of pushing the ball once, he will now forcefully push itover and over again until he is reinforced. It might have been different with adifferent betta, however, and if so, you could follow the same process ofgradually working up to these more difficult criteria.

There was one thing I was very conscious of and did my bestto avoid when training Duo to push the ball, and that was excessive circlingaround the ball before moving to bite it. When I trained this behavior with apast betta of mine, Puck, I found that he would always circle the ball severaltimes before pushing it, which was not really conducive to the end behavior of fetchingthe ball. This was a natural thing for him to do in a situation like this (if you stick your finger inthe water, for example, you will often see a betta circle it several timesbefore they move to bite it) when working himself up to biting the ball, and Ihad unintentionally reinforced him so that he learned to circle the ball everytime before pushing it. With Duo, I was very careful to shape the behavior in away that did not reinforce circling the ball before pushing it,and have not had the same difficulty with him that I had with Puck. I imaginethis would be important for a lot of bettas, which is why I mention it here.

Moving forward, I will start to reinforce Duo only when hepushes the ball in my direction, and then eventually, push it away and thenreinforce him for pushing it back to me (starting with very small distances and then working up to longer ones). Eventually, he should be able to fetch the ball and push it back to meregardless of where I initially put in his aquarium.

Something else to note that occurred with Duo was hisfrustration early on in the training process, as he would often choose to stopparticipating in the training session and would swim laps around his aquariumflaring instead. This is why I don’t have early videos (even though I wanted torecord the process visually, as it is easier for some people to understand ifthey can actually see what I am talking about); as soon as he saw the camera,he would leave and start flaring instead. I really had to pay attention to hisbody language, and either stop working on this behavior to switch to something thatwas easier for him, or make sure toend the training session on a positive note before he chose to leave on hisown. I had to be spot-on with the timing of my reinforcement, and also spent alot of time on each step, really making sure he was responding positivelybefore moving on to the next. It’s a non-issue now and he is very enthusiasticabout working with the ball (and doesn’t mind being recorded!), but it’s something I’ll still have to keep in mindmoving forward.

That’s about all I can think of at the moment! I hope itwas helpful in some way! Training Duo has been a lot of fun and a great way to get to know him a little better- I highly suggest everyone do some sort of training with their own fishes. It’s a very rewarding process!