The Fungal Microbiome Associated with Schizaphis graminum Biotypes E and K in Response to Host Plant Species, Cultivar, and Elapsed Time

Because so little is known about fungal microbiomes associated with aphids, the fungal mycobiota associated with feeding by biotypes E and K of greenbug aphid (Schizaphis graminum) was surveyed using ITS amplicon sequences with regard to host plant species, days post infestation, host plant cultivar, and host plant susceptibility to greenbug feeding. The host plant species were wheat (Triticum aestivum L.), rye (Secale cereale L.), barley (Hordeum vulgare L.), sorghum (Sorghum bicolor (L.) Moench), and Aegilops triuncialis L., and there were 17 cultivars or accessions among these species. Biotype, collection date, host plant species, host plant cultivar, and host plant susceptibility ranked strongest to weakest in influence on fungal community composition, and all influences were statistically significant except susceptibility, but biotype had by far the greatest influence. Filamentous ascomycetes and yeast-like basidiomycetes dominated the diverse fungal microbiota, and many reads mapped to unidentified taxa at the phylum or class level. Prominent genera of filamentous fungi included Fusarium, Mortierella, Simplicillium, Arthrobotrys, and “Capnodiales genus incertae sedis”. Prominent yeast genera included Hannaella, Papiliotrema, Naganishia, and Rhodotorula among basidiomycetes and “Debaryomycetaceae genus incertae sedis” among ascomycetes. Sister samples of each combination of biotype, cultivar, and collection date could differ greatly in frequency of dominant genera, and individual genera frequently bloomed to high frequency with biotype K. Community diversity decreased at the end of the time course. Analysis of frequency correlations with weighted gene co-expression network analysis (WGCNA) identified six modules of 10 to 28 correlated genera in biotype K and five modules in biotype E. At least five genera in a “Hannaella” module have documented antifungal activity. The sample collection failed to separate greenbugs from adhering honeydew or fragments of the leaf where they were feeding, and therefore the location of the detected fungi is not known, which limits the inferences that can be drawn from correlations of taxon frequencies. Most importantly, there is still no proof that a gut microbiota exists in greenbug. Nevertheless, numerous negative correlations in taxon frequency implied that more than one fungal community was present. Further investigation with microscopy is needed to locate fungi in or on the greenbug or at its feeding site.