The causes and consequences of a diverse and dynamic microbiome

Symbiosis with heritable microbes is now recognized as a widespread phenomenon, especially among the insects, and can serve as adaptive novelty. With respect to the host, symbionts are categorized as obligate, necessary for normal growth and reproduction, or facultative which are not always present in all individuals within a population. Facultative symbionts may be maintained in natural populations as parasites through reproductive manipulation or by providing a beneficial trait to the host. Symbiont communities can also be highly diverse with multiple facultative symbionts often inhabiting the same host. Positive and negative symbiont associations have been observed in natural populations suggesting mutualistic or antagonistic interactions between symbionts or host may regulate symbiont spread and community structure. Variation in symbiont community structure has been observed over broad geographic scales and the course of several years but rarely have these differences been observed on finer, seasonal time scales. We would expect the dynamics of heritable symbionts to follow similar patterns of nuclear inherited alleles that can spread within and across populations due to selection acting on a favored phenotype. Although several cases suggest this is true, in-depth exploration of the factors that promote or constrain spread of facultative mutualists in natural populations is limited. Significant temporal shifts of symbionts in natural populations imply that benefits to the host are context dependent, being realized under some environmental conditions but not others. Pinpointing the specific agents of selection that act to regulate symbiont dynamics in the wild is often tricky and requires controlled field experimentation and extensive sampling over time. To better understand how natural enemies affect symbiont spread in natural insect populations I tracked pea aphid, Acyrthosiphon pisum, facultative symbionts along with their natural enemies. Controlled field and laboratory experiments were conducted to measure the spread of the parasitoid defending symbiont Hamiltonella defensa in the presence or absence of the parasitoid wasp Aphidius erv and assess fitness effects of two H. defensa strains. Lastly, the symbiont communities of aphids from across the United States were assessed along with genotyping of H. defensa. This allowed me to determine the stability of common positive or negative symbiont associations over a broad geographic range and assess symbiont associations with different H. defensa strain variants. Symbionts proved to be highly dynamic, the majority undergoing significant seasonal frequency shifts. Although we found a signal of enemies driving seasonal symbiont fluctuations in one population, this correlative finding was not consistent across multiple populations or over two years within the same population. Aphids surviving during periods of high parasitoid-induced mortality were no more likely to possess Hamiltonella than during periods of low parasitoid-induced mortality. Field cage experiments failed to show spread of Hamiltonella when in the presence of parasitoid wasps but frequencies were higher in field cages with parasitoids compared to field cages that excluded parasitoids in one of two years. Several positive and negative associations between symbiont species were evident across multiple years and populations, in some cases uniting symbionts implicated in distinct functions which should in theory broaden the ecological repertoire of the host aphids. Distinct H. defensa genotypes varied in their tendencies to associate with different symbiont species and H. defensa strain variation over time may have resulted in frequency shifts of some of these symbionts within two populations. In summary, my findings reveal the pea aphid possesses a highly diverse and dynamic symbiont community, yet targeting the mechanisms that maintain and shape that diversity is not so easily understood under field conditions. Field surveys and controlled experiments do not strongly support parasitoid-driven balancing selection as a primary explanation for the maintenance of facultative symbionts in natural populations. Controlled field experimentation will be necessary to elucidate the mechanisms promoting or limiting spread of symbiont communities and critical if we are to better understand tripartite host-symbiont-enemy interactions, even for well-developed lab models like the pea aphid.