The ability to curb natural behaviors in an aquarium is not an easy task and mainly involving patience. This is; however, really only needed for problematic behaviors often seen in angels (large and dwarf) and pufferfish; although, I'm sure it can be used in others such as butterflyfishes. The quarantine process is usually the best time to start, since that is the period when you will be observing specific behaviors along with treatment protocols. Basically, it is your personal time with the animal. Unless it is a difficult feeding animal or inlvolves specific treatment (de-worming), prepared foods are best. Since we cannot exactly replicate what an animal consumes in the wild, I begin by feeding a variety of foods separately to see what the fish prefers to eat, what it is willing to eat, and what it dislikes. Only a select few that are of preference are used to curb coral appetites and the rest fed as a regular diet (in a mix). Normally once your 30 days of normal quarantine (actual treatments included) is over I introduce a specific coral to the animals liking in the wild.
Note: This cannot be done if copper or other harsh medications have been used so the fish must be moved into an untreated holding.
Once the fish begins veering towards the coral, about to bite, I slowly slide my hand in the tank between the two. The fish will naturally swim away. After the fifth repetition in a row (this may take up to 60mins) I offer the fish one of its favorite/preferenced foods as a treat and remove the coral: Reward System. It would be best to work with your fish every day for a month, switching out a genus of coral weekly or bi-weekly. What the fish eventually reacts to, is not seeing coral as a main food source. The fish begins wanting prepared food and not caring about coral content. Call me crazy? Sure, but it has worked for me. The only drawback is the personality of your animal and its food intake, whether it is hungry or just bored. For example, I've had a Centropyge eblii that would begin picking at feather dusters, zoanthids, and sponge if I did not offer a small sheet of nori at least every other day. If it received its favorite food regularly then it showed no interested in living animals. I currently have a pufferish at work (I have pictures on this board if you Search) that I trained the same way. It will consume hard and soft corals, but the reason being is that it was bored. In order to cure its boredom I used food, once again, and made what i call "cookie-cubes" which are a weeks worth of its favorite foods mashed together and frozen in ice cube trays (lasts about a month). If given a half cube twice daily the puffer leaves the coral alone. Boredom seems to be the biggest problem for many fish rather than actually consuming because of hunger, but usually stimulating their appetite keeps them from entering a bored state. It's a double-edged sword of sorts. Your best bet with specific fish you are worried about is not to introduce a singular point of interest. In other words, if you add a single coral it will be the new point of interest and may inspect out of curiosity. I've been, what I call successful, with this approach with Centropyge loricula, Centropyge bispinosa, and Centropyge eibli along with Arothron nigropunctatus, Diodon holacanthus, Canthigaster valentini, Canthigaster jactator, Holacanthus ciliaris, Pomacanthus arcuatus, and Chaetodontoplus mesoleucus. I might be missing some...
Oh yes, for what i think is rather humorous, fish seem to be wary of fingers (index for instance) and can be used in training. Scoff if you want, but I've been able to keep fish not normally fathomable in reef tanks.