Excellent! You've completed the "marking" phase of your mark and recapture. We will exclude the 3 recaptures from the count of 72. What you have are 69 marked fish in your population. All of the recaptures were GSF and there is definitely the potential if you were catching and releasing in the next phase to count a fish twice. You DON'T want to do that. The next survey needs to be comprised only of unique fish (none caught twice or more). Ideally your second survey would not be biased to catchability. An example would be an electrosurvey. Another example would be traps (though perhaps not as unbiased as an electro survey) In any event, even an electro survey might be biased to one species or another. Anyways, I recommend you give your fish a rest for a week or so and then try to collect another group of samples. If you can, try to collect around 70 if you can. Don't just throw them back. Either cull them, or hold them in a trap until the second survey is complete so that you are only counting unique catches. Your initial survey is biased to catchability on earthworms. That's probably a good trait to keep, so fish you have caught twice by that survey technique are probably good ones to keep from a fishability/population management perspective. Anyways, right now, I would be reluctant to cull fish I have caught twice.
On your second survey, at least measure and weigh unique catches (the unmarked fish). Decide if you will ultimately release the unmarked fish back to your population. If the decision is yes, mark it the same as you did those yesterday and put it in the holding pen for release later. For those unmarked that you will cull, chill them in ice until you can later clean them for cooking. I would keep the previously marked fish in the holding pen along with the newly marked fish you are keeping. The number caught in this second survey and the number of recaptures in this second survey (as they relate to the initial survey) will be used to estimate populations.
Your initial survey provides a lot of useful context. If it is reasonable that fishing with earthworms is not biased to the species, then we can use the proportions of species in the catch as reflective of the populations. Your initial survey suggests that PKS are your dominant species and comprise in the neighborhood of 50% of your adult prey fish population. The remaining 50% are split between GSF and YP where the GSF have only a modest numerical advantage in the survey. All of the fish in your initial survey are adult fish that are capable of reproduction. Up to now we don't have any information for potential recruits. My sense (right now) is that recruitment is low.
Of your existing prey fish species, the PKS have the best condition averaging 103 RW, where GSF follow with 86 RW, and YP with 78 RW. All of the different species have comparable average weights PKS (73.57 g), GSF (73.57 g), with YP (79.93 g) possessing the highest average weight. For PKS and YP, RW was negatively correlated with length (RW worsens at longer lengths). This is normal when food is limiting for the population. The GSF, however, have a positive correlation with length. If I had to venture a guess as to why, my guess would be that they have some advantage for consuming some of the prey in the pond. For example, perhaps longer GSF with larger gapes have an advantage with fish and crayfish prey items. The PKS having an overall species advantage for condition, perhaps this is a benefit of exploiting snails in their diet.
The good news. You do not have stunted populations of panfish (although they are admittedly growing slower than they could be growing). I base this on Swingle's definition of a harvest sized panfish. A panfish > 1/10 of pound is considered harvestable. In grams this is around 45.45 grams. >92% of your initial survey are comprised of harvest sized panfish (per Swingle). To be stable, there needs to be a good balance of predator and prey. Presently, the risks to this population structure are losses of the adult panfish population cocurring with an inadequate population of predators (both number and size). A reduced adult panfish population will result in reduced cannibalization of YOY and higher recruitment. This higher survival resulting from a lower adult panfish population needs to be balanced by the control of the right number and size of predators. To only reduce the number of adult panfish would probably result in fewer harvestable size panfish and more smaller panfish. I do think you will need to bump the number of LMB. I like 8"-10" LMB for this purpose. Probably fewer if we are able to pin a good number for your less than 8" LMB.
Did you happen to mark your LMB? Going forward, also do that and keep track of how many have been marked. Always release your LMB, at least for now. The single LMB has good RW (120). This fish is probably in its second year of life ... posssibly later. In population dynamics, LMB less than 8" are not counted. At least not in texts authored by Anderson and/or Willis. I think whether we should count them as part of your solution depends on how fast we can expect them to grow into the 8-12 class of fish. When you have good balance, there will need to be representation in this size.