I think the best way to read the control is to subtract the control's gain from the gain of the two treatments. This will give a true conversion of the feed. But the problem is that you have lots of minnows in the pond that the free roaming fish can slaughter. It may be possible that the fish that roam free will outgain both of your cages. To isolate the effect of the feed's contribution ... one has to limit the control's access to pond food organisms in the same way the fed fish will be limited. This can only happen if the control fish are caged ... equal in number and weight ... in the same sized cage. If control fish are caged and do not gain ... then you have evidence that ALL of the gain -could- be attributable to feed. There is still time if you have fish in the right size and number.
Even so, some residual bias due to wastes possibly attracting pond food organisms could remain. In other words, feeding may attract natural foods preferentially to the fed cages relative to the control cage but this would be a whole different experiment
We can only speculate whether there is any effect
