Frank, so I think your thoughts could be graduated from a notion to an idea. It is a testable hypothesis. It would qualify as knowledge if tested against a control experiment where the effect could be quantified in terms of percent survival. I think you could just do it anyway perhaps fattening a fin clipped female F1 for release into your pond each cycle.

Fish are pretty good at learning from their fallen comrades.

Last year, I re-established a population of GAMs in a small pond with abundant LMB. The pond had once harbored GAMs but the introduction of LMB did a number on them and I was unable to observe any around the margins of the pond.

So a pond nearby hosted GAMs in profuse numbers. No LMB in that pond but many BG, GSF, and CC. Once the GAMs got to 1.5" they seemed to have little fear of the existing predators and they swam wherever they pleased ... with impunity. Not so of the small GAMs which are very skittish jumping clear of the water as one would walk around the ponds edge. So it would seem that a fish can also learn when it is safe as well. Now big GAMs aren't entirely without fear. They don't like to have anything out of water above them and they would keep a distance from me that seemed to be appropriate to evade GBH. They preferred water too deep for GBH to wade and so I would have to net them around two concrete structures that afforded water that was more than 6 feet deep.

In netting the GAMs I found they quickly connected my actions with predation and with each swipe of my net they became much more difficult to capture. Soon whey would keep a distance that was just enough to be out of reach of my handle. They retained this behavior even on my next outing. So I took an old broom stick and I extended the reach. Now this worked very well, but soon thereafter they adopted behavior that kept them just greater than my reach with the then extended net. I recall wondering if they knew the must keep more distance when I was with the net than without it. Anyways, I was impressed with how fast they learned to evade me.

When this GAMs were released (June & July), their existing behaviors were in full play and true to form the big pregnant GAMs swam to water too deep for GBH. But .... they were met with LMB for the very first time ... who promptly stalked and ate them. But they were not dumb GAMs, they recognized the threat and tried to evade LMB leaping from the water. Still many succumbed. So I wonder if would make a difference. I only stocked maybe 400 GAMs many of which were consumed same day but they reproduced and there were thousands by Fall. This spring, my impression was that no more than 200 or 300 survived winter. Somehow enough survived to contribute to food chain. This year, I have already noticed that the first hatch of GAMs is around 3/8 to 1/2 inches long in the middle of May. For posterity I think I might stock 300 or 400 adult females this year and see if it makes a difference this Fall with an additional cohort of young earlier in the season.

Other approaches to increase survival is to time culling prior to releases and to design the forage pond to accommodate 24 hours of loss. There is a limit to the quantity of prey an LMB can eat. It is some proportion of their weight. So if they could eat 12% of the their weight and the standing weight of LMB is 80 lbs/acre then that transferred that exceeds 10 lbs/acre should learn to fear LMB