Flowers as viral hot spots: Honey bees (Apis mellifera) unevenly deposit viruses across plant species

Abstract

RNA viruses, once considered specific to honey bees, are suspected of spilling over from managed bees into wild pollinators; however, transmission routes are largely unknown. A widely accepted yet untested hypothesis states that flowers serve as bridges in the transmission of viruses between bees. Here, using a series of controlled experiments with captive bee colonies, we examined the role of flowers in bee virus transmission. We first examined if honey bees deposit viruses on flowers and whether bumble bees become infected after visiting contaminated flowers. We then examined whether plant species differ in their propensity to harbor viruses and if bee visitation rates increase the likelihood of virus deposition on flowers. Our experiment demonstrated, for the first time, that honey bees deposit viruses on flowers. However, the two viruses we examined, black queen cell virus (BQCV) and deformed wing virus (DWV), were not equally distributed across plant species, suggesting that differences in floral traits, virus ecology and/or foraging behavior may mediate the likelihood of deposition. Bumble bees did not become infected after visiting flowers previously visited by honey bees suggesting that transmission via flowers may be a rare occurrence and contingent on multiplicative factors and probabilities such as infectivity of virus strain across bee species, immunocompetence, virus virulence, virus load, and the probability a bumble bee will contact a virus particle on a flower. Our study is among the first to experimentally examine the role of flowers in bee virus transmission and uncovers promising avenues for future research.

Fig 1. Schematic of experimental designs.

In a series of four experiments, we examined virus deposition on flowers by honey bees and/or virus transmission between honey bees and bumble bees. To examine the effect of plant species and/or plant diversity, flowering plant species were provided to foraging bees as either single plant species (A) or in diverse arrays consisting of all three species (B). To test whether chronic exposure to contaminated flowers is necessary for virus transmission, bumble bee microcolonies were exposed three times to honey bee-visited flowers over the course of three days (C). To test if direct contact or comingling is necessary for viral transmission, honey bees and bumble bees were allowed to forage together in tent enclosures (D). Blue boxes in the schematic represent tent enclosures assigned as the honey bee tent (where infected honey bees were allowed to forage on flowers), the exposed bumble bee tent (where plants exposed to honey bees were transported into three hoop houses to be foraged on by bumble bee microcolonies), and the control bumble bee tent (where bumble bee microcolonies foraged on ‘clean’ plants brought directly from the greenhouse). Red semi-circles represent hoop houses within bumble bee tents, each containing a single bumble bee microcolony. Green arrows represent the movement of plants from the honey bee tent to the exposed bumble bee tent after a 15 hour nectar regeneration period. In the chronic experiment, the same three bumble bee colonies were used on each of three days (depicted by red arrows connecting the hoop house through time). Three plant species were used throughout the series of experiments: Trifolium repens, T. pratense, and Lotus corniculatus. Photos of inflorescences and tent enclosures (with hoop houses) are provided for visualization (E).

Fig 2. Virus load for virus positive flower samples by plant species across all trials.

Box plots color coded by plant species. Whiskers represent max and min, the box edges are the 1st and 3rd quartiles and the middle line represents the median. Deformed wing virus (DWV), black queen cell virus (BQCV). Plant species are Lotus corniculatus (birdsfoot trefoil), Trifolium pratense (red clover), or Trifolium repens (white clover).