This project has accumulated a lot of observations of visibly parasitized caterpillars. Some caterpillar species have hundreds of qualifying observations.

So now that we have this mass of observations, what can we do with them? One simple thing is to get a feel of rates of parasitism among different caterpillar species. Just compare the number of parasitized caterpillars in this project to the total number of caterpillars, species by species.

Here's that math for most-reported caterpillars in the project:

Carolina Sphinx (Manduca sexta)
386 parasitized of 6017 total = 6.4% of caterpillars visibly parasitized

Cecropia Moth (Hyalophora cecropia)
115 of 1690 = 6.8%

Regal Moth (Citheronia regalis)
124 of 767 = 16.2% (yikes!)

Saddleback Caterpillar Moth (Acharia stimulea)
70 of 1952 = 3.6%

Spiny Oak-slug Moth (Euclea delphinii)
51 of 661 = 7.7%

Of course, there are some potential issues with this methodology. Below is a quick list of some conceivable criticisms. Are any of these likely to change the rates significantly? Are there others that I'm overlooking?

• For my counts of parasitized and total reported caterpillars, I am looking only at observations where the intended subject was the caterpillar. So I could be overlooking observations of parasitized caterpillars that were submitted as observations of the parasite. Also, for simplicity, I am counting each observation as a single caterpillar, regardless of how many caterpillars are photographed.
• Are parasitized caterpillars significantly under- or over-reported on iNat? Maybe parasitized caterpillars are more likely to be photographed, either due to impaired concealment, or because they might arouse more interest in the observer? Or is the opposite true; parasitized individuals might become less active, and therefore less noticeable?
• One source suggested that a particular tachinid fly parasitizes only final-instar caterpillars. So photos of early instars don't necessarily represent individuals that will never be parasitized; they haven't yet reached the stage where they could be parasitized. If this is true then the percentages above are underestimates. (And in general, this is probably true even if parasitism isn't restricted to the final instar.) But maybe the percentages above are still meaningful in a relative sense, and you could still accurately make a statement like, "E. delphinii caterpillars are twice as likely to be parasitized as A. stimulea caterpillars."

And of course I am explicitly not addressing parasitism that happens in other life stages, like while pupating.

A big thanks to @claire146963 and many others for annotating the life stages in these thousands of observations.

Doug