I find this work to be compelling and the study is a rigorous test of self-medication. I especially like the focus on adaptive plasticity (as the authors say, it “empowers the science of self-medication” – but I wonder if others in the self-medication field think that it needed to be empowered in this way?). I completely agree that associative learning is not a required component of self medication and that the adaptive plasticity context is a better way to think about this phenomenon.
I look forward to seeing how general self-medication really is in caterpillars, since I’m a little biased towards thinking that parasites in general may be well adapted to toxins in their hosts’ food plants.

These questions arose as I read this publication:
• What is the biological significance of 16-18% changes in survival and mortality? I wonder if this would have any population-level consequences?

• I am particularly interested in the mechanism of modifying parasitism. How are the ingested PAs toxic (or otherwise detrimental) to tachinids? It would be interesting to see if there is any genetic variation in E. mella’s ability to detoxify or resist PAs.

• The statistics seem fine here, but there are so many statistical problems with choice tests that complex patterns like those found here are difficult to interpret (see the vast literature on this; e.g., Lockwood 1998, Oecologia 116:475-481; Manly 1993, Oecologia 93:149-152; Roa 1992, Oecologia 89:509-515). In contrast, the results from the no-choice test were easy to interpret. This is a pretty common problem, so I think the best way to proceed is to combine no-choice tests, pairwise tests, and cafeteria tests (which in this case would involve greater variation in available PA diets).

• It was not clear from this paper how many eggs are usually oviposited by tachinids on (or near) arctiids in general, but I would not ignore the possibility that some time in time or space these caterpillars have been attacked by tachinids that get 3 or more eggs or larvae, so I’m not sure it is that important from a pure theoretical point of view to characterize this as unusual or extreme.

• I think that understanding how encapsulation is affected by PAs and other aspects of Grammia’s diet will help clean up some error variance here, and it should contribute to understanding competing demands of utilizing diet to enhance the immune response while also defending against predators and parasites.