Consequences of symbiont co-infections for insect host phenotypes.

Most animals host communities of symbiotic bacteria. In insects, these symbionts may have particularly intimate interactions with their hosts: many are intracellular and can play important roles in host ecology and evolution, including protection against natural enemies. We investigated how interactions between different species or strains of endosymbiotic bacteria within an aphid host influence the outcome of symbiosis for both symbiont and host. We first asked whether different combinations of facultative symbiont species or strains can exist in stable co-infections. We then investigated whether the benefits that facultative bacteria confer on their hosts (protection against natural enemies) are enhanced, reduced or unaltered by the presence of a co-infecting symbiont. We asked this both for co-infecting symbionts that confer different phenotypes on their hosts (protection against fungal pathogens vs. parasitoid wasps) and symbionts with overlapping functions. Finally, we investigated the additional survival costs to aphids of carrying multiple infections of symbiont species or strains, and compared symbiont titres in double and single infections. We found that stable co-infections were possible between all of the combinations of facultative symbiont species (Regiella insecticola + Hamiltonella defensa, Regiella + Rickettsiella sp., Regiella + Spiroplasma sp.) and strains (Hamiltonella) that we studied. Where symbionts provided protection against different natural enemies, no alteration in protection was observed in the presence of co-infections. Where symbionts provided protection against the same natural enemy, the level of protection corresponded to the higher of the two symbionts present. In some instances, aphid hosts suffered additional survival costs when hosting double infections. In the case of Hamiltonella, however, infection with multiple strains of the same symbiont led to lower symbiont titres than single infections, and actually improved aphid survival. We conclude that the long-term maintenance of symbiont co-infections in aphids is likely to be determined primarily by costs of co-infections and in some instances by redundancy of symbiont benefits.