Molecular identification of Cerodontha australis (Diptera: Agromyzidae) and its associated pupal endoparasitoids (Hymenoptera)

Abstract

The wheat sheath miner, Cerodontha australis, is widespread and abundant in New Zealand and also occurs in eastern Australia. Adult and larval C. australis feed on cereals and grasses, including the economically important perennial ryegrass, Lolium perenne. There is little literature about C. australis, and much of the early work may relate to other species due to its initial misidentification. Morphology-based identification can be challenging, and the absence of online barcode sequences from voucher specimens for C. australis has precluded diagnosis using molecular techniques. In this study, two individual adult leafminers collected from L. perenne plants in New Zealand were confirmed as C. australis morphologically. One has been retained as a voucher specimen and its mitochondrial barcode sequence submitted to online databases. Comparison with adult and larval specimens previously sequenced by the authors confirmed they were also C. australis. Molecular identification of 20 hymenopteran pupal endoparasitoids, and a subset of the puparia they emerged from, revealed Trichopria sp. (Diapriidae), Eupelmus messene (Eupelmidae) and three haplotypes (or closely related species) of Eulophidae. Sequences for all wasps were submitted to GenBank. All the puparia were visually identified as C. australis, and a subset that included a host pupa of each identified endoparasitoid species were confirmed as C. australis based on comparison with our barcode sequence. This work will aid future studies on C. australis and has allowed identification of unidentified sequences on public databases formerly submitted by other New Zealand researchers. It also provides new records of association between C. australis and its pupal endoparasitoids. More broadly, our study provides an example of the importance of authoritative morphological identification of specimens alongside molecular identification, and highlights the challenges in identifying species for which this has not been done.