Search

Facultative intracellular symbionts like the α-proteobacteria Wolbachia influence their insect host phenotype but little is known about how much they affect their host microbiota. Here, we quantified the impact of Wolbachia infection on the bacterial community of the cabbage root fly Delia radicum by comparing the microbiota of Wolbachia-free and infected adult flies of both sexes. We used high-throughput DNA sequencing (Illumina MiSeq, 16S rRNA, V5-V7 region) and performed a community and a network analysis. In both sexes, Wolbachia infection significantly decreased the diversity of D. radicum bacterial communities and modified their structure and composition by reducing abundance in some taxa but increasing it in others. Infection by Wolbachia was negatively correlated to 8 bacteria genera (Erwinia was the most impacted), and positively correlated to Providencia and Serratia. We suggest that Wolbachia might antagonize Erwinia for being entomopathogenic (and potentially intracellular), but would favor Providencia and Serratia because they might protect the host against chemical plant defenses. Although they might seem prisoners in a cell, endocellular symbionts can impact the whole microbiota of their host, hence its extended phenotype, which provides them with a way to interact with the outside world.

A major challenge in network ecology is to describe the full-range of species interactions in a community to create highly-resolved food-webs. We developed a molecular approach based on DNA full barcoding and mini-barcoding to describe difficult to observe plant-leaf miner-parasitoid interactions, consisting of animals commonly regarded as agricultural pests and their natural enemies. We tested the ability of universal primers to amplify the remaining DNA inside leaf miner mines after the emergence of the insect. We compared the results of a) morphological identification of adult specimens; b) identification based on the shape of the mines; c) the COI Mini-barcode (130 bp) and d) the COI full barcode (658 bp) fragments to accurately identify the leaf-miner species. We used the molecular approach to build and analyse a tri-partite ecological network of plant-leaf miner-parasitoid interactions. We were able to detect the DNA of leaf-mining insects within their feeding mines on a range of host plants using mini-barcoding primers: 6% for the leaves collected empty and 33% success after we observed the emergence of the leaf miner. We suggest that the low amplification success of leaf mines collected empty was mainly due to the time since the adult emerged and discuss methodological improvements. Nevertheless our approach provided new species-interaction data for the ecological network. We found that the 130 bp fragment is variable enough to identify all the species included in this study. Both COI fragments reveal that some leaf miner species could be composed of cryptic species. The network built using the molecular approach was more accurate in describing tri-partite interactions compared with traditional approaches based on morphological criteria.