Author: Derek Dunn
Position: Postdoctora Research Associate
Institution: University of Reading, UK
E-mail: d.w.dunn@reading.ac.uk
Additional Authors: S.T. Segar; R. Chan; J. Ridley; D. Yu, R. Crozier & J.M. Cook
Submitted Date: April 08, 2008
Published Date: April 9, 2008
This comment was originally posted as a “Reader Response” on the publication date indicated above. All Reader Responses are now available as comments.

We welcome the comments from Pereira on our recent paper (Dunn et al. doi:10. 371/journal.pbio.0060059) and agree that our results invite new detailed investigations of the effects of non-pollinating fig wasps on the fig-pollinator mutualism. There are over 750 Ficus species with many of the sub-generic sections (clades), endemic to particular biogeographical regions. Wider understanding of mutualism stability will require detailed studies of how parasites, pollinators and trees interact in these different lineages on different continents.
We are aware of the paper by Anstett (1), which we cited for the hypothesis that inner flowers nurture larger female pollinators than outer flowers. However, our data did not support this prediction. We have since explored Anstett's ideas, and formulated our own, on how flower position influences early emergence from the syconium. Using the same species of Ficus as in our PloS paper, we showed that female wasps in the inner flowers were not more likely to emerge early from the syconium, but that females in outer galls did suffer a greater risk of remaining trapped in their galls (female wasps in F. rubiginosa cannot leave their ovules unless males release them (2)). These results contrast with Anstett's (1) data for F. microcarpa; although we note that she measured emergence from galls into the syconium, while we studied emergence from the syconium into the external world. We think that biological mechanisms causing foundresses to favour laying eggs in inner flowers are likely to vary between Ficus species, and that factors affecting mutualism stability are also likely to vary across Ficus (3,4), again calling for comparative studies.
Rates of parasitism vary between Ficus populations (e.g. considerably in our study) and species, but one has to look quite hard to find odd examples of natural populations of any endophytic insects (in general)that are not subject to parasitoid attack. However, the impact of parasitism is bound to vary between Ficus species.
Pereira emphasises how little is known about the larval biology of most non-pollinating fig wasps and that larval diet may be far more labile than is often assumed. We echo this and other calls for dedicated studies on this subject (4), which is important not only to mutualism stability but also for characterisation of fig wasp food webs. Interestingly, larval diets and food webs are better understood for hymenopteran parasites of another radiation of endophytes – cynipid wasps that develop in oak galls (5) - and some species appear to be facultative, potentially feeding on plant tissue, or insect tissue, or both. It would be surprising if non-pollinating fig wasps do not include such cases, as well as hyperparasites. However, we end by noting that externally ovipositing species with facultative diets could still place selection on pollinators to avoid outer flowers if one amongst their options is to usurp a pollinator gall and kill the occupant.

1. Anstett 2001. Oikos. 95: 476-484.
2. Dunn et al. 2008. Journal of Animal Ecology. In press.
3. Herre 1999. In Keller L. (ed) Levels of Selection in Evolution. PUP. Princeton.
4. Cook & Rasplus 2003. Trends Ecol Evol. 18: 241-248.
5. Stone et al. 2002. Annu. Rev. Entomol. 47: 633-668.
6. Yu et al. 2004. Proc. Roy. Soc. B. 271: 1185-1195