The analysis of interaction networks across spatial environmental gradients is a powerful approach to investigate the responses of communities to global change. Using a combination of DNA metabarcoding and traditional molecular methods we built bipartite Drosophila – parasitoid food webs from six Australian rainforest sites across gradients spanning 850 m in elevation and 5°C in mean temperature. Our cost‐effective hierarchical approach to network reconstruction separated the determination of host frequencies from the detection and quantification of interactions. The food webs comprised 5–9 host and 5–11 parasitoid species at each site, and showed a lower incidence of parasitism at high elevation. Despite considerable turnover in the relative abundance of host Drosophila species, and contrary to some previous results, we did not detect significant changes to fundamental metrics of network structure including nestedness and specialisation with elevation. Advances in community ecology depend on data from a combination of methodological approaches. It is therefore especially valuable to develop model study systems for sets of closely‐interacting species that are diverse enough to be representative, yet still amenable to field and laboratory experiments.

Background Kissing bugs (Triatominae) are blood-feeding insects best known as the vectors of Trypanosoma cruzi, the causative agent of Chagas’ disease. Considering the high epidemiological relevance of these vectors, their biology and bacterial symbiosis remains surprisingly understudied. While previous investigations revealed generally low individual complexity but high among-individual variability of the triatomine microbiomes, any consistent microbiome determinants have not yet been identified across multiple Triatominae species. Methods To obtain a more comprehensive view of triatomine microbiomes, we investigated the host-microbiome relationship of five Triatoma species sampled from white-throated woodrat (Neotoma albigula) nests in multiple locations across the USA. We applied optimised 16S rRNA gene metabarcoding with a novel 18S rRNA gene blocking primer to a set of 170 T. cruzi-negative individuals across all six instars. Results Triatomine gut microbiome composition is strongly influenced by three principal factors: ontogeny, species identity, and the environment. The microbiomes are characterised by significant loss in bacterial diversity throughout ontogenetic development. First instars possess the highest bacterial diversity while adult microbiomes are routinely dominated by a single taxon. Primarily, the bacterial genus Dietzia dominates late-stage nymphs and adults of T. rubida, T. protracta, and T. lecticularia but is not present in the phylogenetically more distant T. gerstaeckeri and T. sanguisuga. Species-specific microbiome composition, particularly pronounced in early instars, is further modulated by locality-specific effects. In addition, pathogenic bacteria of the genus Bartonella, acquired from the vertebrate hosts, are an abundant component of Triatoma microbiomes. Conclusion Our study is the first to demonstrate deterministic patterns in microbiome composition among all life stages and multiple Triatoma species. We hypothesise that triatomine microbiome assemblages are produced by species- and life stage-dependent uptake of environmental bacteria and multiple indirect transmission strategies that promote bacterial transfer between individuals. Altogether, our study highlights the complexity of Triatominae symbiosis with bacteria and warrant further investigation to understand microbiome function in these important vectors.