Flowers Are Spreading Viruses Among Wild Bees

More flowers lead to more pathogen spread between bee species.

More flowers should mean healthier bees, right? Not always. Scientists studying wild bees in Israel have discovered that greater flower diversity can sometimes increase viral infections rather than reduce them, complicating a common assumption about how biodiversity protects against disease.

The finding adds nuance to the belief that diverse plant communities naturally dilute pathogen spread by reducing contact between species. Instead, researchers found that in certain contexts, a richer variety of flowers creates new pathways for viruses to jump between different bee species.

Unexpected Link Between Variety and Virus Spread

The research team surveyed 14 sites across Israel’s Judean Foothills during spring 2018, screening 153 wild mining bees and 150 honey bees for four common viruses. Mining bees are solitary species that typically specialize on just a few types of flowers. At most sites in the study, these bees visited yellow-flowering mustard plants almost exclusively (about 99% of their flower visits).

However, in areas with higher flower diversity, something else happened. Sites with more flower varieties showed increased rates of deformed wing virus (DWV) in mining bees. The virus prevalence climbed alongside both the number of flower species and the density of honey bees foraging in the area.

The likely culprit? A flower called Erucaria hispanica. This plant appeared to act as a bridge species, attracting both the normally specialist mining bees and generalist foragers like honey bees. When diverse flowers pulled different bee species together onto shared blooms, viruses had new opportunities to spread.

The study, published in Ecological Applications, also found that sites with more flower varieties attracted more bee visitors to the yellow mustard plants that mining bees depend on. Rather than spreading out across many flower types, increased competition appeared to concentrate different bee species onto the same plants.

Certain Flowers Protect While Others Facilitate Disease

While overall diversity showed concerning patterns, individual flower species played different roles. Two plants (Echium judaeum and Notobasis syriaca) were associated with lower virus prevalence for multiple viruses. In contrast, sites with more Erucaria hispanica showed higher infection rates for both black queen cell virus and Lake Sinai virus-2.

These differences may stem from flower traits that affect virus transmission. Flowers with open, exposed structures could intercept more sunlight, which may degrade pathogens and reduce infection risk for visiting bees. Flower shape also influences how bees interact with blooms, and consequently how viruses get deposited or picked up.

Landscape-Scale Flower Availability Amplifies Infection Risk

The findings extend well beyond individual gardens or farm fields. Researchers calculated how many flowers were available across the broader landscape surrounding each sampling site, looking at areas up to 1,000 meters away, or more than half a mile.

Higher flower availability at these landscape scales correlated with increased virus prevalence for three of the four viruses studied: black queen cell virus, Lake Sinai virus-2, and sacbrood virus. The effect was stronger at the landscape level than at individual sites.

This landscape connection suggests that far-ranging honey bees, which can forage over several miles, may spread viruses across considerable distances. Even mining bees with more limited flight ranges can encounter viruses introduced from the broader landscape, possibly through occasional long-distance dispersers or via shared flowers visited by mobile honey bees.

Honey Bees Carry Viruses to Wild Species

Every site where researchers detected viruses in mining bees also had virus-positive honey bees. Genetic analysis provided direct evidence for at least one virus: sacbrood virus strains from honey bees and mining bees showed 98.3% identical genetic sequences, indicating recent transmission between species.

For sacbrood virus, mining bee infection rates climbed with the density of infected honey bees foraging at a site. For deformed wing virus, total honey bee density mattered more than specifically infected individuals. Thus, honey bees might influence mining bee vulnerability through resource competition and nutritional stress rather than just direct virus transfer.

Virus prevalence in honey bees ran much higher than in mining bees across all four viruses tested, ranging from 20% to 86% in honey bees compared to 5% to 12% in mining bees. This difference reflects the fundamental distinction between social and solitary lifestyles. Honey bees live in dense colonies where constant contact and food sharing create conditions for rapid disease spread, while solitary mining bees have limited contact with others of their species.

What This Means for Pollinator Conservation

The findings carry implications for conservation efforts aimed at supporting pollinators through flower plantings, particularly in Mediterranean agricultural systems during spring bloom periods. Creating diverse wildflower meadows has become a popular strategy for helping bees, and for good reason, as pollinators need abundant and varied food resources. Unfortunately this research suggests those habitats could also create conditions for increased disease transmission under certain circumstances.

The relationship between diversity and disease depends heavily on context: which specific plant species are present, what other bee species are foraging in the area, and how the landscape is structured at larger scales. Rather than concluding that diverse plantings are harmful, the research points to the need for more nuanced approaches. Land managers might consider which specific plant combinations attract different bee species, potentially reducing harmful overlap.

The landscape-scale effects also matter. Pollinator health initiatives need to look beyond individual habitat patches to consider how flower resources are distributed across entire regions.

Researchers selected 14 sites with flower diversity ranging from just one species to 14 species during a two-week period in mid-March 2018 at peak spring bloom. The mining bee sample included four closely related Andrena species with similar foraging behaviors. Virus detection used RNA extraction followed by PCR to identify the four aforementioned viruses.

---