Also remember that the article states (in the abstract): "Genetic analyses indicate that tadpoles are frequently unrelated to embryos in the same pool" - this may indicate that the same male won't deposit into a pool already used.No, I wasn't thinking they'd egg feed. I was thinking of the intentional cannibalism of R. variabilis , and the fact that tads in non-egg feeders may well be deposited in an occupied canister, replicating the effect of the natural behavior of variabilis. I presume that while the vanzolinii group knows enough to stop breeding when caring for eggs, the fantastica group might well not (no adaptive reason to), and will thus end up feeding their earlier tads. But no, I wasn't planning to count on that, and plan to provide food.
Interesting that this could also indicate that such strategies could have been what developed species like imitator or flavovittata into facultative egg feeding. I don't remember where the species diverged though, so I can't comment on that at present.