Periodic Report Summary 1 - SYMBIOCONTROL (The role of the bacterial symbiont Cardinum in shaping the ecology and evolution of an insect parasitoid wasp: fundamental studies and implications for biological control of plant pests.) Introduction. Speciation, the splitting of single evolutionary lineages into reproductively isolated groups, remains one of the most elusive phenomena in the study of evolution. Reproductive isolation between populations is a landmark event in the speciation process because, by restricting further gene flow, all subsequent genetic changes occur independently in different populations that are diversifying. Heritable bacterial symbionts play a role in causing or reinforcing reproductive isolation. Maternally inherited bacteria that cause reproductive incompatibility (cytoplasmic incompatibility or CI) in their arthropod hosts are among the most prevalent. Incompatibility occurs when infected male hosts mate with uninfected females, producing few or no viable offspring. CI symbionts spread because of their ability to depress the fitness of uninfected females relative to infected ones. Two bacteria have been shown to induce CI. Wolbachia (α-Proteobacteria) is the best-known reproductive manipulator, infecting as many as 66% of arthropods. More recently, Cardinium (Bacteroidetes) has been found to cause CI in the parasitoid wasps Encarsia pergandiella, Encarsia inaron and in two spider mites. Since its discovery, Cardinium has been found in about 7% of arthropods. CI-inducing bacteria may contribute, along with other mechanisms of isolation, to overall isolation between populations. While it has been established that Wolbachia can play a role in speciation, the role of Cardinium is unknown. Similarly, the cytological mechanism of CI is known only for Wolbachia. Invasive species threaten agricultural and natural systems throughout the world, and often thrive at the expense of indigenous, closely related organisms. Biological control of invasive species with natural enemies from the region of origin has been shown to be one of the most effective and environmentally sustainable forms of management. However, when populations with reproductive barriers and/or different CI bacterial infection status come into contact in the release area, reproductive failure may occur, and one or both populations may be suppressed. Nevertheless, relatively few studies have examined the implications of symbiont infection status for biological control efficacy. Hymenopteran parasitoids (wasps that complete development at the expense of a single insect host, killing it in the process) are highly specialized enemies of herbivorous insects, and are commonly infected with a wide range of facultative symbionts. Encarsia pergandiella is a tiny (0.6-0.7 mm) parasitoid of whiteflies that has been used in the biological control of worldwide insect pests of field and glasshouse crops, like Bemisia tabaci and Trialeurodes vaporariorum. Encarsia pergandiella from California (hereafter CA) is uninfected by symbionts, while E. pergandiella from South Texas (hereafter TX) is infected by a Cardinium that causes CI. Despite being separate species, CA and TX are still able to hybridize. Among the reproductive barriers that occur in the process of speciation, the incompatibility arising from hybridization (hybrid incompatibility or HI) can also interfere with the effectiveness of biological control. Cardinium-induced CI and HI may hence have important implications for the practical use of E. pergandiella in biological control. Aims and scope. Using Cardinium and E. pergandiella as endosymbiont and insect models, respectively, the project SYMBIOCONTROL aims to: 1) evaluate the role of CI-inducing bacteria and genetic isolation mechanisms in reproductive isolation and competitive interactions of parasitoid populations; 2) evaluate how the introduction of a new biological control agent that is partially reproductively isolated (via CI or HI) from an established parasitoid population affects the biological control of a key agricultural pest. During the outgoing phase, we have addressed objective 1. Encarsia pergandiella represents an ideal model to disentangle the effects of CI and HI because the infected TX species can be cured from its Cardinium infection by antibiotic treatment (we refer to the cured strain as cTX).Methods and results. We first carried out a challenging taxonomic revision of the E. pergandiella species complex. By integrating molecular, morphological and morphometric data, we showed that this complex includes five species rather than two, three of which we described for the first time. These species can be now distinguished not only by biological but also by molecular markers and morphological features. Then we assessed the extent of pre-zygotic and post-zygotic isolation between the CA and TX species. We tested for sexual isolation by performing no choice, female choice, male choice and multiple choice mating experiments, recording crossing outcome and courtship, copulation and competition behaviour. While no choice tests resulted in mating in all interspecific treatments, female, male and multiple choice tests resulted in overall significant sexual isolation, with significant mating asymmetries between the two species. We assessed the relative contribution of CI and HI to the post-zygotic isolation by performing all possible nine crosses (including controls) among CA, TX and cTX in small arenas. We found a different mortality phenotype associated with CI and HI. By scoring these two phenotypes, we showed that both CI and HI play a role in the reproductive isolation, with CI having a significant additive effect in the interspecific CI-incompatible cross, whereas HI was stronger in the other direction. F1 hybrid progeny and the few F2 progeny obtained were mostly sterile, suggesting that HI is almost complete at this stage of the speciation process. In order to assess the effect of reproductive incompatibility, in particular if and how CI and HI contribute to competition in a secondary contact scenario, we performed population cage experiments between CA-TX, and CA-cTX. As it is very difficult to discriminate between parental type wasps and hybrids on the basis of morphology and body colour, we developed a highly effective molecular typing method based on a High Resolution Melting protocol. Cage experiments showed that, at present time, due to the accumulation of HI, the effect of CI in the competition is negligible, whereas mating preferences play an important role. The results of small arenas and cage experiments will be used to build a theoretical framework for the effect of CI on speciation in this system. Finally, we elucidated the cytological mechanism behind the CI phenotype. By performing staining of live and fixed eggs of E. pergandiella, and observing the eggs using confocal and deconvolution microscopes, we showed that the Cardinum symbiont is able to manipulate the reproduction of its hosts by interfering with the cleavage divisions at the early embryogenesis stage. When uninfected eggs were fertilized by infected males, several anomalies were recorded starting from the first mitotic division up to the formation of the embryo, such as the formation of chromatin bridges, different degrees of chromatin condensation, deformed nuclei and micronuclei, resulting in the death of embryos.The impact. These results provide an important advancement in the knowledge of the role of symbionts causing unidirectional CI in promoting speciation, which had been so far only shown for the closely related species Drosophila recens and D. subquinaria. The Drosophila/Wolbachia and Encarsia/Cardinium systems each present peculiarities that are critical for a better understanding of speciation, one of the most elusive and fascinating questions in evolutionary biology. Furthermore, understanding the role of heritable microbial symbionts in the biology of natural enemies has not yet been widely considered as an important research objective of biological control programmes. Many authors have proposed that biological control needs to become more predictive, yet despite considerable discussion on the topic, few biological control programs have attempted to evaluate prospective biocontrol agents in a predictive manner. If the predictive methods used in a biological control program are accurate, the time needed to identify and evaluate an effective natural enemy can be shortened, which has value in reducing the impact of the target pest. By elucidating the effect of CI and HI on the isolation and competition among closely related species of biocontrol agents, SYMBIOCONTROL will provide a theoretical and practical framework for future fundamental and applied studies.
