Many parasitoids in the families Braconidae and Ichneumonidae have obligatory mutualistic relationships with bracoviruses and ichnoviruses, respectively. Interestingly, while these viruses have similar functional relationships with their respective parasitoid families, these have been independently acquired and represent clear examples of convergent evolution.

This study explores the possibility of convergent evolution in the way that these endosymbiotic viruses play important roles in mediating the tritrophic relationships between plants, caterpillar herbivores, and their parasitoids. The authors use a study system (Brassica oleracea, Pieris brassicae, and its two parasitoids – the braconid Cotesia glomerata and the ichneumonid Hyposoter ebeninus) to which they have made many important contributions.

The manuscript is clearly written. The experiments are clearly described and the results for the most part are straightforward. The overall aim of this study is to explore whether both the bracovirus and ichnovirus both alter the HIPV profiles such that the parasitized hosts are less attractive to conspecific parasitoids. The authors provide evidence suggesting that both viruses (in the oral secretions of the caterpillars) alter plant volatile profiles such that these caterpillars are less attractive to superparasitizing wasps, arguing that this represents an example of convergent evolution.

I was puzzled by the decision to not study the effects of venom injected by Hyposoter ebeninus. This is a serious deficiency in my mind. The authors state that this was because the venom a congener, H. didymator has no discernable role in parasitism success in a study by Dorémus and colleagues (2013). However, this is a different species of parasitoid, and the host caterpillar, Spodoptera frugiperda, is in a very different family! (Noctuidae vs. Pieridae), making inferences to the current study system problematic. Furthermore, S. frugiperda has recently invaded Europe (and elsewhere) from the Americas, suggesting that the relationship between H. didymator (and its ichnovirus) and S. frugiperda has not co-evolved. This study would be far stronger if the authors conduct an additional experiment where they study the combination of HeIV+Venom (similar to what they did with CgBV+V). This is especially relevant as in the C. glomerata system, bracovirus or venom alone had no effect on parasitoid preference, and it was only the combination that induced a change in HIPVs. Maybe your conclusions would be similar to what is currently written, but it would be far stronger.

Other questions:
I was curious why you used different approaches for examining preference of the two parasitoid species: wind tunnels for C. glomerata and a static two-chamber olfactometer for H. ebeninus.

Please give the g-force of your rpm measures (or at least the radius of your centrifuge) (line 182 and elsewhere). It would be helpful in the event that someone wishes to replicate your study.

Line 268-270: You use t-tests to compare the proteomes of the P. brassicae treatments. Do you use any sort of adjustments (e.g., Bonferroni or similar) as you are running multiple tests?

Starting at line 297, you state that you found 126 proteins that differ between the regurgitant of the PBS caterpillars (unparasitized) vs. the Cg-parasitized or CgBV+V treatments and 122 different proteins in the similar H. ebeninus study. I was curious how much overlap there was between these two different groups – or were they completely different?

It’s interesting that you found no deterrence of heterospecific wasps, but how common is it that you find multiparasitized hosts in the field?

Finally, could you speculate (I realize that the Discussion is already on the long side – maybe there are places where you could be more concise) about why you see an effect of HeIV alone on the foraging behavior of H. ebeninus (figure 4) but no effect of CgBV alone on the foraging decisions of C. glomerata (figure 3)? It’s only when CgBV is combined with venom do you see an effect on the foraging decisions of C. glomerata. Again, it would be nice to see data from an experiment where you included a HeIV+V treatment. Given that H. ebeninus is the stronger competitor during events of multiparasitism, one might expect that C. glomerata would avoid hosts parasitized by H. ebeninus. It’s interesting that you did not find this to be the case, but I wonder whether a HeIV+V vs. PBS treatment would have resulted in a preference for PBS for C. glomerata? Again, I feel that additional experiments are warranted.

I appreciate the detailed and thorough responses that the authors provided for my previous queries (as well as those of the other reviewer). These responses satisfactorily address all my concerns apart from one (and even this is partially addressed as explained below).

One of my previous concerns focused on the decision to exclude treatments involving venom of Hyposoter ebeninus. I appreciate the more detailed list of studies/reviews that contend that ichnoviruses do not play a significant role in synergizing the activity of IV in ichneumonid parasitoids. I agree that this makes it less likely (although not inconceivable given the limited number of ichneumonid study systems that have explored this question) that you would expect to see an effect of an HeIV + venom treatment in H. ebeninus and P. brassicae. More convincing would be a study that focused on this particular parasitoid – caterpillar association. At the very least, an HeIV + venom treatment would serve as a control treatment that would confirm that He-venom plays no role in the activation of HeIV against P. brassicae.

This (albeit attenuated) criticism aside, I felt the authors did a nice job of addressing the other comments of the previous version.