Parasitoids Turn Herbivores into Mutualists in a Nursery System Involving Active Pollination.

Nursery pollination involves pollinators that lay eggs on the flowers they pollinate and have their brood fed on flower parts or developing ovules [1-4]. Active pollination, a ritualistic behavioral sequence shown by nursery pollinators when transferring pollen from anthers to stigmas, is known in only four plant lineages [5-8], including the classical examples of fig trees-fig wasps and yuccas-yucca moths [5, 6]. We report in detail a system in which weevils actively pollinate orchids prior to having their larvae fed on the developing fruits. Sampling over five years revealed that although weevils trigger fruit set, this interaction is negative for the plant as weevil larvae often consume all contents of infested fruits. However, part of weevil-infested fruits is often "rescued" by parasitoid wasps, which kill the weevil larvae before all fruit content is consumed (Figure 1). "Rescued" fruits present high seed viability and biomass similar to that of non-infested fruits, much higher than that of fruits with weevils only. Hence, parasitoids mediate the fitness consequences of the interaction between the plant and its parasitic pollinator. Weevils constitute a megadiverse group of herbivores commonly reported as florivores [9] but are also appreciated as flower-ovipositing pollinators of cycads and palms [4, 10-13] and were previously recorded carrying orchid pollinaria [14-16]. The orchid-weevil system presented here shows that plant-floral visitor interaction outcome can be mediated by a third party (parasitoids) and illustrates a way by which the biological context may allow the emergence and persistence of active nursery pollination behavior in nature.

Laboratory rearing of solitary bees and wasps.

Ecological experiments often require standardized methods that exclude natural variation and allow manipulation of a single parameter. It has been shown that domesticated honey bee larvae are raisable in a controlled environment. Here we demonstrate that this approach is also transferable to wild solitary bees and wasps without inducing negative effects on their development. Wells may also be supplemented with the antibiotic substance oxytetracycline to control the presence of bacteria. The method thus provides a useful tool to investigate offspring recruitment and larval development in solitary bees and wasps, plus their responses to manipulation of factors as for example diets, toxins and microbiota.

Prophenoloxidase from Pieris rapae: gene cloning, activity, and transcription in response to venom/calyx fluid from the endoparasitoid wasp Cotesia glomerata.

Prophenoloxidase (PPO) plays an important role in melanization, necessary for defense against intruding parasitoids. Parasitoids have evolved to inject maternal virulence factors into the host hemocoel to suppress hemolymph melanization for the successful development of their progeny. In this study, the full-length complementary DNA (cDNA) of a Pieris rapae PPO was cloned. Its cDNA contained a 2 076-base pair (bp) open reading frame (ORF) encoding 691 amino acids (aa). Two putative copper-binding sites, a proteolytic activation site, three conserved hemocyanin domains, and a thiol ester motif were found in the deduced amino acid sequence. According to both multiple alignment and phylogenetic analysis, P. rapae PPO gene cloned here is a member of the lepidopteran PPO-2 family. Injection of Cotesia glomerata venom or calyx fluid resulted in reduction of P. rapae hemolymph phenoloxidase activity, demonstrating the ability to inhibit the host's melanization. Real-time reverse transcriptase polymerase chain reaction (RT-PCR) showed that transcripts of P. rapae PPO-2 in the haemocytes from larvae had not significantly changed following venom injection, suggesting that the regulation of PPO messenger RNA (mRNA) expression by venom was not employed by C. glomerata to cause failure of melanization in parasitized host. While decreased P. rapae PPO-2 gene expression was observed in the haemocytes after calyx fluid injection, no detectable transcriptional change was induced by parasitization, indicating that transcriptional down-regulation of PPO by calyx fluid might play a minor role involved in inhibiting the host's melanization.

Population biology of cytoplasmic incompatibility: maintenance and spread of Cardinium symbionts in a parasitic wasp.

Bacteria that cause cytoplasmic incompatibility (CI) are perhaps the most widespread parasites of arthropods. CI symbionts cause reproductive failure when infected males mate with females that are either uninfected or infected with a different, incompatible strain. Until recently, CI was known to be caused only by the alpha-proteobacterium Wolbachia. Here we present the first study of the population biology of Cardinium, a recently discovered symbiont in the Bacteroidetes that causes CI in the parasitic wasp Encarsia pergandiella (Hymenoptera: Aphelinidae). Cardinium occurs at high frequency ( approximately 92%) in the field. Using wasps that were recently collected in the field, we measured parameters that are crucial for understanding how CI spreads and is maintained in its host. CI Cardinium exhibits near-perfect rates of maternal transmission, causes a strong reduction in viable offspring in incompatible crosses, and induces a high fecundity cost, with infected females producing 18% fewer offspring in the first 4 days of reproduction. We found no evidence for paternal transmission or horizontal transmission of CI Cardinium through parasitism of an infected conspecific. No evidence for cryptic parthenogenesis in infected females was found, nor was sex allocation influenced by infection. We incorporated our laboratory estimates into a model of CI dynamics. The model predicts a high stable equilibrium, similar to what we observed in the field. Interestingly, our model also predicts a high threshold frequency of CI invasion (20% for males and 24% for females), below which the infection is expected to be lost. We consider how this threshold may be overcome, focusing in particular on the sensitivity of CI models to fecundity costs. Overall our results suggest that the factors governing the dynamics of CI Wolbachia and Cardinium are strikingly similar.