Microbes are easily dispersed from one place to another, and immigrant microbes might contain information about the environments from which they came. We hypothesized that part of the microbial community on a flower's surface is transferred there from insect body surfaces and that this community can provide information to identify potential pollinator insects of that plant. We collected insect samples from the field, and found that an insect individual harbored an average of 12.2 × 105 microbial cells on its surface. A laboratory experiment showed that the microbial community composition on a flower surface changed after contact with an insect, suggesting that microbes are transferred from the insect to the flower. Comparison of the microbial fingerprint approach and direct visual observation under field condition suggested that the microbial community on a flower surface could to some extent indicate the structure of plant–pollinator interactions. In conclusion, species-specific insect microbial communities specific to insect species can be transferred from an insect body to a flower surface, and these microbes can serve as a “fingerprint” of the insect species, especially for large-bodied insects. Dispersal of microbes is a ubiquitous phenomenon that has unexpected and novel applications in many fields and disciplines.

1. The flower visitor community consists not only of pollinators but also of non‐pollinators, such as florivores, thieves and predators that attack flower visitors. Although there is increasing evidence that early‐season foliar herbivory influences pollinator visitation through changes in floral traits, few studies have explored indirect effects of foliar herbivory on community structure of the flower visitors. We examined how early‐season foliar herbivory influences the flower visitor community established in late season. 2. We conducted an inoculation experiment using a lacebug (Corythucha marmorata), which is a predominantly herbivorous insect attacking leaves of tall goldenrod (Solidago altissima) in Japan. 3. Flower abundance significantly decreased when damaged by the lacebug. The numbers of pollinators, florivores and thieves were positively correlated with flower abundance, whereas predators were not. In response to flower abundance, florivores decreased on damaged plants. On the other hand, thieves increased on damaged plants, and pollinators and predators did not differ between damaged and undamaged plants. 4. When effects of flower abundance were excluded, foliar herbivory still influenced florivores negatively and thieves positively. This implies that factors besides flower abundance may have affected the numbers of florivores and thieves. 5. Community composition of flower visitors on damaged plants significantly differed from undamaged plants, although overall abundance, taxonomic richness and taxonomic evenness were unaffected by foliar herbivory in the early season. It is important to recognise that only evaluating species diversity and overall abundance may fail to detect the significant consequence of early‐season herbivory on the flower visitor community.

No abstract

No abstract