Recovering ecological interactions by mining non-target data from whole genome re-sequencing projects

The study of parasitic species can shed light on aspects of their hosts ecology. Such interactions are however often unknown or understudied due to the difficulty to detect and/or quantify many infections. Whole-genome sequencing and re-sequencing data have been generated at an increasing rate and reduced costs over the last two decades. Projects based on whole organisms, like whole insect specimens, provide genomic material for the target taxon, but may also include sequencing reads from associated microbes and other parasites. Here, we screened for the presence of non-host reads in a collection of whole-genome (re-) sequence projects from the Glanville fritillary butterfly, Melitaea cinxia, a model organism in research on the ecology and evolution of species in spatially structured and fragmented landscapes. We identified infections with different bacteria and eukaryotic parasites, which are shared between populations and underly both previously known and new biotic interactions for this butterfly species. The bacterial symbiont Wolbachia, usually common in insects, was found at relatively low prevalence, while Spiroplasma was ubiquitous across samples from several European populations of the butterfly. Additionally, we confirmed expected rates of larval parasitism by at least two parasitoid wasps. Such results provide proof of principle that key ecological interactions can be uncovered efficiently from existing data, an important first step to characterising the role host-associated organisms play in shaping the ecology and evolutionary history of their host species.