Identification of adult vectors of Xylella fastidiosa (Xf) based on morphological characters can be difficult for non-specialists. The identification of eggs or larvae is even more difficult and sometimes impossible. To circumvent these difficulties and enable large survey of populations of insect vectors in Europe we have developed a web-interface database of COI barcodes. COI barcodes were generated with a high-throughput approach involving two PCR steps to target COI and index samples followed by Illumina sequencing. Multiple quality controls were implemented including i) non-destructive DNA extraction that allows vouchering and morphological re-examination by taxonomists; ii) bioinformatic controls to discards possible contaminants, PCR or sequencing errors, NUMTs and heteroplasms. All barcodes have been uploaded in a web-interface database (Arthemis DB@se: http://arthemisdb.supagro.inra.fr/DefaultInfo.aspx?Page=Home). Metadata associated with all records are available online, as well as biological data on vectors and pictures. Users can identify a query sequence through BLAST comparison against our reference library. Currently, the database hosts 636 barcodes for 72 species of potential vectors of Xf: 490 sequences representing 43 European species and 145 sequences representing 29 non-European species. All frequent species have been barcoded. Missing species are rare or localized and do not represent an important threat for European agriculture. COI allows reliable identification for 80% of the species, but we observed issues for the remaining 20%. The most important issue concerns Philaenus spumarius and its close relative P. tesselatus. These two species are hardly distinguishable even with male genitalia and share the same COI sequence. It was also difficult to cross validate morphological and molecular results for some species within the genera Neophilaenus and Aphrophora.

FR; PPT; jean-claude.streito@inrae.fr