I can't see how eliminating competition among your own would lead to being a predator on a different species. I think that most interesting is if there is some method to where the moth lays its eggs, say to increase the chances of finding a snail (doesn't say if any old snail will do or if only certain kinds are prey). As you know ichneumons lay their eggs directly in(on) their prey. That certainly means that some will be successful. Perhaps the reason is in the lack of insect predators in Hawaii, so they can be slow and allow for many "misses" and still be a successful species.
Keep in mind that snails are voracious plant-eaters too. Most gardeners find snails/slugs a more destructive pest than caterpillars. It's not a big leap from eating others of your kind when you encounter them as "competition" on your hostplant, to eating *other* invertebrates you find on the same plant for the same reason. I'd bet a reasonable sum of money that the snails are common "predators" on the same hostplant that this caterpillar's ancestors used to feed on. The snails may have been bigger "competition" for food than others of the caterpillar's own kind.
Eventually it may have turned out to be more productive to take a step up the food chain and just specialize on eating the snails exclusively.
Or maybe not exclusively. The article doesn't mention whether the caterpillars eat snails *exclusively*, or hunt them in *addition* to still eating the ancestral hostplant. From the description, it doesn't sound as if they've been studied in the wild much. That may still be an open question. And it seems a *long* evolutionary road from being an obligate vegetarian to an exclusive carnivore, even among insects. Such a drastic change in diet requires a lot of biochemical adjustments. Mixed diets are a lot easier to adjust to, since you can still get essential nutrients from your ancestral foodsource, and don't have to "switch" over to getting *all* of them from a completely different kind of diet.
In any case, a scenario like the one above seems a much more plausible evolutionary pathway than just starting from a plant-eating ancestor, and then *bam*, becoming snail-hunters in "one step".
As for how caterpillars started eating others of their own kind, when they were "built" for eating plants, some have hypothesized that it's not that big an innovation, since other eaters of the same hostplant may well "smell" or "taste" like the hostplant which the caterpillars are primed to chemically recognize and consume anyway. Most caterpillars are highly selective about their hostplants, and seem to have specific chemical "triggers" which tell them which plants are suitable hostplants and which ones aren't. Another caterpillar (or snail?) which has been eating the same hostplant may well be exuding the same chemicals which it has incorporated from its own meals. Caterpillars aren't all that bright, and anything which "smells" the same may be considered an acceptable meal, even if it's not green and leafy.
I saw a possible instance of this the other day. Monarchs aren't normally cannibalistic, and large dense batches of them (denser than would ever be found in nature) can be raised on the same milkweed without noticeable mortality -- they just climb over each other. But the other day I saw one Monarch caterpillar eat the "feeler" off another one, in exactly the same manner it would have eaten a sliver of milkweed. I think it was just keying off the same alkaloids it uses to recognize milkweed itself, which Monarch caterpillars concentrate in their own bodies in order to make them distasteful to predators like birds.
Nor have I ever seen a Monarch (or any other caterpillar) "experimentally" eat anything other than its appropriate hostplant, so that explanation for the "feeler eating" is unlikely. Caterpillars will starve to death rather than eat (or even attempt to eat) a plant not on their "genetically approved" list. If it doesn't have the right chemical triggers, it just ain't food, as far as they're concerned. We lost two batches of Spicebush Swallowtail caterpillars until we realized that what was sold to us as a spicebush plant by a usually reliable nursery wasn't actually spicebush. The caterpillars all just failed to eat, and died. Even some hybrid varieties of their actual hostplant will be refused by caterpillars, apparently they don't "smell" right. This is a common problem with Pipevine Swallowtails -- many hybrid pipevine varieties favored by gardeners are rejected by the caterpillars, which is a real shame because the *adult* Pipevine Swallowtails often recognize them as suitable hostplants and lay eggs on them, but the *caterpillars* don't, and subsequently die of starvation. Apparently the adults and the caterpillars use somewhat different "recognition" systems -- not surprising, considering their difference in physiology. But the end result is that a lot of Pipevine Swallowtail eggs get "wasted", which may be one of the reasons that they are rather rare, especially compared to the success stories of the other swallowtail species.
But I digress. My point is that while caterpillars would be expected to make "chemical recognition mistakes" and occasionally eat other consumers of their preferred hostplant, natural selection would likely only encourage serious cannibalism in species where hostplants are more likely to be totally consumed by overly large broods. This could explain why some species are frequent cannibals and others aren't. And a similar mechanism could easily have led this particular moth species into a situation where it was eating snails in the first place (due to chemical miscues caused by eating the same hostplant), and was also under selective pressure to wipe out as many snails as it could (if the snails frequently caused the caterpillar's hostplant to end up denuded of leaves before the caterpillars could finish their fifth instar and pupate).
I can't think of a more plausible evolutionary scenario (heck, I can't think of *any* scenario) by which the caterpillars would by some other means make the leap from eating plants in their ancestral form, to going after snails in a serious way. It seems especially unlikely for caterpillars to begin eating snails by some other scenario when you consider just how highly selective they are in their eating habits -- if they don't even eat other plants that don't "smell" right, even hybrid plants of the same genus, how would they start going after snails by any scenario *other* than the sort of "tastes the same due to food competition" scenario mentioned above?
I think that most interesting is if there is some method to where the moth lays its eggs, say to increase the chances of finding a snail (doesn't say if any old snail will do or if only certain kinds are prey).
See above -- if the snails frequent the same ancestral hostplant, the adults don't even need to change their egg-laying habits. Laying on the same hostplant places the eggs in proximity to the snails which the caterpillars now eat.
As you know ichneumons lay their eggs directly in(on) their prey. That certainly means that some will be successful. Perhaps the reason is in the lack of insect predators in Hawaii, so they can be slow and allow for many "misses" and still be a successful species.
True -- insects have a general reproductive strategy of making up in quantity what they might lack in individual success rates. Butterflies and moths typically lay upwards of a hundred eggs per female. They can afford to lose a few and use a "spray and pray" methodology.
