The evolutionary ecology of an insect-bacterial mutualism Thesis submitted in accordance with the requirements of the University of Liverpool for the degree of Doctor of Philosophy By Chris Corbin September 2017 1 Abstract Heritable bacterial endosymbionts are responsible for much phenotypic diversity in insects. Mutualists drive large-scale processes such as niche invasion, speciation and mass resistance to natural enemies. However, to persist, mutualists need to be able to transmit with high fidelity from one generation to the next, to be able to express their beneficial phenotypes, and for the benefits they grant the host to outweigh their costs. The effect of ecologically-relevant environmental temperature variations upon transmission and phenotype is a poorly understood area of endosymbiont biology, as is how the symbiont’s cost varies under ecological stress. In this thesis, I examined these parameters for Spiroplasma strain hy1, a defensive mutualist which protects the cosmopolitan, temperate fruit fly Drosophila hydei from attack by a parasitoid wasp. I detected Spiroplasma hy1 in D. hydei individuals from the south of the U.K. The bacterium is at low prevalence compared to hy1 in other localities such as North America and Japan, but its presence in this temperate region conflicts with past studies indicating high sensitivity to low temperatures. I first demonstrate that the vertical transmission of Spiroplasma hy1 is more robust to the cool temperatures typical of temperate breeding seasons than previously considered, with transmission in a ‘permissive passage’ experiment occurring at high fidelity for two generations at a constant 18°C and in an alternating 18/15°C condition. Secondly, I demonstrate that the expression of the defensive phenotype is considerably more sensitive to cool temperatures than transmission. Spiroplasma hy1 protection ceases at 18°C, suggesting that for much of the D. hydei breeding season in areas such as the U.K., hy1 may be selectively neutral in many fly individuals. Finally, I show that hy1 has an unusually low standing cost to its host under starvation stress, contrasting with findings for the related MSRO strain in D. melanogaster. Measures of active cost – the fate of survivors of attack – were unclear. These results indicate that sensitivity to cold temperatures could account for hy1’s low U.K. prevalence. Small amounts of segregational loss could partially counteract selection upon natural enemy resistance, and loss of phenotypic expression at 18°C almost certainly causes hy1 to be neutral at best for parts of early summer and autumn. Future work should investigate the effects of different temperature on costs of symbiont carriage, and whether cool temperatures could push hy1 from mutualism and neutral commensalism to parasitism, as well as investigate how nuclear-mediated anti-wasp protection might interact and compete with hy1-mediated protection. 2 Acknowledgments Firstly, I would like to thank my supervisors, host institution, and funders. In addition to interesting science discussions and excellent guidance, Greg Hurst provided vast quantities of snacks, and had good humour and patience in the face of insect stock collapses, his students printing him fake degree certificates, and having his office filled with balloons. Andy Fenton provided valuable help with using R for statistics. Thanks also go to the University of Liverpool for hosting me, and to NERC for funding my PhD. I was lucky to receive help from many sources through the course of my project. Thank you to Darren Obbard, who provided abundant Drosophila hydei flies from Tunbridge Wells for two years. Ben Longdon caught specimens of D. hydei from Cambridge. Fabrice Vavre gave me L. heterotoma wasps, from which I set up my laboratory stock. Rowan Connell was a source of invaluable support, helping with starvation assays and the PCR workload when I was carrying out the passive cost experiment. Gabriel Pedra and Amanda Minter taught me more R and helped me work through statistics problems. I’d like to thank my coauthors on the review paper published as part of this thesis, Greg Hurst, Ellie Heyworth, and Julia Ferrari, as well as two anonymous reviewers. Additionally, my thanks go to the members of the EE and EEGID groups at the University of Liverpool, and to other members of the Hurst lab, who often provided feedback on my presented work and ideas (and babysat my flies when I went on holiday). On a personal note, I’d like to thank my fellow PhD students, but most especially Daria Pastok, Amanda Minter, and Sarah Trinder. They gave me much-needed sensible advice, co- constructed elaborate cakes, and took part in a variety of impromptu lunch break diversions, including sewing and knitting sessions and a post-apocalyptic science fiction book club. Thanks again to Rowan Connell, who is both relentlessly positive and fed my pets when I was away at conferences or on holiday. I’d like to thank Kyle Lyon for making me cups of tea and being a listening ear when science plans weren’t working out. Thank you to Riverside Rebels Roller Derby for teaching me to roller skate, taking me on camping adventures, and for being the best group of friends anyone could ask for. Finally, a great big thank you to my pet rats, despite their apparent complete lack of interest in my research. 3 Table of contents Abstract .................................................................................................................................... 2 Acknowledgments .................................................................................................................... 3 Table of contents ..................................................................................................................... 4 List of figures ............................................................................................................................ 9 List of tables ............................................................................................................................. 9 1 Introduction ........................................................................................................................ 10 1.1 The majority of arthropods carry bacterial endosymbionts ........................................ 10 1.2 Endosymbionts underlie a range of unusual phenotypes in arthropods which don't otherwise make sense ....................................................................................................... 11 1.3 Symbionts have impacts on the evolutionary and ecological dynamics of arthropods ........................................................................................................................................... 12 1.3.1 Sex ratio skewing, and resulting counter-adaptations to reproductive parasites 12 1.3.2 Genetic sequences from reproductive parasites can insert into genomes, with consequences including the production of new sex determination systems................ 12 1.3.3 Protective mutualists add further complexity to host-parasite dynamics ........... 13 1.3.4 Symbionts of all classes are capable of driving speciation and cladogenesis ....... 14 1.3.5 A host species may evolve dependency on its symbiont and thus be constrained by its needs .................................................................................................................... 15 1.3.6 Horizontal transmission of a symbiont into a novel host represents a mechanism of ‘fast evolution’ in the new host ................................................................................. 15 1.4 Biotic and abiotic factors can change a symbiont’s transmission efficiency and phenotype, causing changes in the population biology of the host-symbiont pair .......... 16 1.5 Temperature may alter several parameters of evolutionary ecology, but is understudied at ecologically-relevant temperatures ........................................................ 17 1.6 Cost of symbiont carriage is of interest as a less widely-studied component of symbiont phenotype and as an impact on symbiont evolutionary ecology ...................... 17 1.7 The study system ......................................................................................................... 18 1.7.1 Spiroplasma strain hy1 is protective against a Drosophila parasitoid, Leptopilina heterotoma .................................................................................................................... 18 1.7.2 A temperate fruit fly, Drosophila hydei ................................................................ 20 1.7.3 Leptopilina heterotoma is a generalist parasitoid on Drosophila and is likely to be a significant selective force for D. hydei ........................................................................ 21 1.7.4 Despite being advantageous against L. heterotoma, Spiroplasma hy1 exists at low to intermediate frequencies in D. hydei ........................................................................ 22 4 1.8 Outline of thesis: what factors could be keeping a 'good mutualist’ down? .............. 23 1.8.1 Chapter 2 – A review of temperature’s influence on heritable symbionts .......... 23 1.8.2 Chapter 3 (part 1) - Is hy1 present in Drosophila hydei in the U.K.? .................... 24 1.8.3 Chapter 3 (part 2) - How is the transmission of hy1 in Drosophila hydei affected by ecologically-relevant low temperature? ................................................................... 24 1.8.4 Chapter 4 - How is the phenotype of hy1 in Drosophila